(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Journal of the Society of Motion Picture Engineers"



From the collection of the 



n 



PreTnger 
v JJibrary 



San Francisco, California 
2007 






JOURNAL 

OF THE SOCIETY OF 

MOTION PICTURE ENGINEERS 

SYLVAN HARRIS, EDITOR 
Volume XVIII JANUARY, 1932 Number 1 

CONTENTS 

Page 

Lighting of Sound Films Louis DUNOYER 3 

The Rapid Record Oscillograph in Sound Picture Studies 

A. M. CURTIS, T. E. SHEA, AND C. H. RUMPEL 39 

Photographic Sensitometry, Part III LOYD A. JONES 54 

Thermionic Tube Control of Theater Lighting . . BURT S. BURKE 90 

A Portable Non-intermittent Cine Projector 101 

Committee Activities: 

Report of the Projection Practice Committee 107 

Report of the Projection Theory Committee 113 

Abstracts 116 

Patent Abstracts 122 

Officers 130 

Committees 131 

Contributors to This Issue 133 

Society Announcements 134 



JOURNAL 

OF THE SOCIETY OF 

MOTION PICTURE ENGINEERS 

SYLVAN HARRIS, EDITOR 



Published monthly at Easton, Pa., by the Society of Motion Picture Engineers 

Publication Office, 20th & Northampton Sts., Easton, Pa. 
General and Editorial Office, 33 West 42nd St., New York, N. Y. 



Copyrighted, 1931, by the Society of Motion Picture Engineers, Inc. 



Subscription to non -members, $12.00 per annum; to members, $9.00 per annum, 
included in their annual membership dues; single copies, $1.50. A discount 
on subscriptions or single copies of 15 per cent is allowed to accredited agencies. 
Order from the Society of Motion Picture Engineers, Inc., 20th and Northampton 
Sts., Easton, Pa., or 33 W. 42nd St., New York, N. Y. 

Papers appearing in this Journal may be reprinted, abstracted, or abridged 
provided credit is given to the Journal of the Society of Motion Picture Engineers 
and to the author, or authors, of the papers in question. 

The Society is not responsible for statements made by authors. 

Entered as second class matter January 15, 1930. at the Post Office at Easton, 
Pa., under the Act of March 3, 1879. 



LIGHTING OF SOUND FILMS* 



LOUIS DUNOYER 



Summary. The author examines different types of illumination apparatus. 
He discusses their conditions of operation and describes t}ie apparatus devised by 
himself in greater detail. This apparatus obtains an extremely fine exploring zone 
simply by projecting the image of a rectilinear incandescent filament on the film 
by means of a good objective. In order to correct the aberrations rigorously, the light 
coming from this filament goes through the walls of the lamp in a place where they 
form parallel faces worked optically. 

In conclusion, the author describes some comparative tests made on apparatus with 
a slit and on the rectilinear filament apparatus. The flux emitted by the latter is 
superior to the flux emitted by the apparatus with the slit, with a consumption ap- 
proximately one-tenth as great. At the same time the disadvantages of the slits (dust, 
defective uniformity of illumination, delicacy of centering, etc.} are eliminated. 

INTRODUCTION 

1 . Review of the Principle of Sound Films. It is well known that 
on a sound film the section set aside for sound production is a small 
straight band only 3 millimeters wide, in general located between 
one of the series of perforations for moving the film and the edges 
of the picture images which are to be projected on the screen. On 
this small band the sounds first have been recorded by either of the 
two following processes which we shall review briefly to make them 
clearer. 

In the constant density** process the sound band is divided in 
two regions, each of which has a uniform photographic density, 
one clear and the other dark, and whose common boundary is a line 
which is more or less indented or wavy. The bends of this line corre- 
spond to the recorded sound vibrations. In most cases this re- 
cording is performed by means of an oscillograph which receives 
the current from the receiving microphone after amplification by 
a triod. In vibrating, the spot of the oscillograph, which consists 
of a small luminous line perpendicular to the length of the band, 

* Translated from Revue d'Optique, 10 (Jan.-Feb., 1931), Nos. 1-2, pp. 1-21, 
57-68. 

** I.e., variable width. 

3 



LOUIS DUNOYER 



[J. S. M. P. E. 



produces an image on the film of a width which varies according 
to the amplitude and the frequency of the sound vibrations. Fig. 1 
shows an example of constant density recording. 

In the variable density process the photographic density of the 
sound band is the same at all points of its width but this density 
varies in the direction of the length of the band. The recording 
system which is used most at the moment consists in letting the 
film slide by a very fine slit illuminated by a flashing lamp. This 
lamp contains a gas under a low pressure illuminated by the dis- 
charge; the voltage at which this discharge is produced is modu- 
lated by the current from the recording 
microphone, conveniently amplified. The 
luminescence of the gas follows these modu- 
lations which are the more intense the 
higher the voltage. These modulations, 
therefore, are transferred to the sound 
band by corresponding modulations of the 
density in the direction of the length of 
the band. Figs. 2 and 3 show two ex- 
amples of variable density recording. Fig. 
2 refers to an invariable musical note (ap- 
proximately 440 vibrations per second). 

In order to reproduce the sounds the 
entire width of the sound band must be 
illuminated, but only for a width equal to 
that of the spot or the slit which illumi- 
nated it during the recording; then having 
passed through the film the light is received 
by a photoelectric cell. If the photographic 

opacities are proportional to the luminous fluxes received during the 
recording, and if the photoelectric currents are proportional to the 
luminous fluxes received by the cell, these currents finally will be 
proportional to the currents of the recording oscillograph or to the 
brilliancy of the flashing lamp, according to the recording process 
employed. Then they are amplified and sent into a loud speaker. 
The distortion of the sound can be due only to the microphone 
circuit or the circuit of the loud speaker. 

Since the light received by the cell should be only the light which 
has passed through the film in a rectangle 3 millimeters long and a 
few hundredths of a millimeter high (0.05 mm. at most) much light 




FIG. 1. An example of 
constant density recording. 



Jan., 1932] 



LIGHTING OF SOUND FILMS 



evidently must be concentrated in this rectangle, and naturally this 
entire flux must fall on the cell after having diverged on leaving 
the film. The latter condition is readily attained; the former con- 
stitutes the problem of illuminating devices for sound films. 

We shall divide this paper in two parts. In Part I we shall first 
review briefly the various types of devices, and then examine the 
theoretical conditions which must be satisfied by the mode of illumi- 
nating the film in order to obtain a suitable sound performance. 




FIG. 2. Example of variable density recording. 

In Part II we shall apply the results of this investigation; we shall 
discuss the properties of existing devices, and describe in detail the 
apparatus which we have devised and the results of a few tests. 

PART I 
THEORETICAL STUDY OF THE LIGHTING OF SOUND FILMS 

2. On the Width of the Illuminated Exploring Region. It is neces- 
sary in particular to know the length on the film occupied either by 
a period of the separating curve in the case of constant density re- 
cording, or by a period of the density in the case of variable density 



6 



LOUIS DUNOYER 



[J. S. M. P. E. 



recording, assuming, of course, that the recorded sound itself has a 
definite period. Let / be the length on the film occupied by one 
period and v the speed of unwinding the film. If the frequency of 
the sound vibration is N, we have: 



'-i 



since the film advances by / in 1/N second. 

In general, the film speed is 45.5 centimeters per second. 



(1) 



One 




FIG. 3. Example of variable density recording; a constant frequency note. 

period of Ia 3 or normal la, corresponding to 440 vibrations per second, 
thus would occupy a height of 1.03 millimeters on the film. It is 
necessary, however, to record much shriller sounds. First, the 
highest note used in music is re-i, corresponding to 4698 vibrations 
per second, which on the film gives a period of 0.097 millimeter. 
But this is far from sufficient for the correct recording of different 
timbres and for the timbre of the human voice. Much higher har- 
monics then must be attained and some hold that it will be necessary 



Jan., 1932] 



LIGHTING OF SOUND FILMS 



to record 20,000 vibrations per second, which on the film would corre- 
spond to a period of 0.02275 millimeter. We shall see the ratio 
which it is possible to allow between the period of the sound and the 
height of the illuminated region on the film. It is clear in any case 
that this height should be smaller than the period. It can be from 
0.2 to 0.05 millimeter. 

3. Different Types of Lighting Apparatus. The majority of ap- 







FIG. 4. One type of lighting apparatus in which the slit 
is placed as near as possible to the film. 

paratus employed heretofore for illuminating the cell through the 
film involves the use of a fine slit which limits the height of the 
exploring zone. This slit can be used in two different ways. 

In one group of devices (Fig. 4), the slit F is placed as near as 
possible to the film. The light supplied by an incandescent lamp, 
with the filament S as concentrated as possible, is projected on 
this slit by means of a condenser. The light which has passed 





FIG. 5. Another type of lighting apparatus in which the image of the slit 
is projected on the film. 

through the film then falls on the photoelectric cell C. Since the 
light rays diverge more from the window the wider the angular 
aperture of the beam from the condenser, thereby increasing the 
illumination of the slit, the illuminated region of the film will be 
sufficiently narrow only if the slit is very close. 

In another group of devices (Fig. 5) the image of the slit is pro- 
jected on the film by means of an objective 0. This objective fre- 



LOUIS DUNOYER 



[J. S. M. P. E. 



quently is a microscope objective which produces a smaller picture 
on the film than the slit itself. The slit, therefore, can be com- 
paratively wide but still all the light received by the condenser is 




FIG. 6. Diagram for studying lighting effects on the film in the 
system illustrated by Fig. 4. 

far from utilized. In this type of device the film is located at a dis- 
tance from the back face of the microscope objective which is equal 
to its frontal distance, that is, a few millimeters. 

In order to avoid the considerable losses of light which take place 




FIG. 7. Similar to Fig. 6, showing a different phase in the passage 
of the film past the aperture. 

in these devices, the extreme nearness of the film and the slit in the 
former, and in a general manner the inconveniences outlined in Part 
II, which result from the use of a slit, devices without a slit can be 



Jan., 1932] LIGHTING OF SOUND FlLMS 9 

used, and of these there also are two types. In one of these types 
a cylindrical or cylindrospherical lens is used. In the other type 
the reduced image of a rectilinear incandescent filament is produced 
on the film by means of a good objective. This solution, which is 
our own, then requires the construction of a special lamp. 

4. Distribution of the Light on the Film. Let EE' (Figs. 6 and 
7) be the exploring zone, or more exactly, its projections on a plane 
through the axis of the sound band and perpendicular to the film, 
this plane being the plane of the figure. Whatever lighting appara- 
tus is used, this plane also is a plane of symmetry of the illuminat- 
ing beam, and the exploring zone receives light from a surface HH' 
which is the exit pupil of the illuminating apparatus; the luminous 
flux passing through a point of the exploring zone fills a cone whose 
peak is this point and the base this pupil. Let FF' be the film; 
for perfect illumination its plane should coincide with EE'. If the 
exploring zone consists of a slit (apparatus of the first type), the 
film should rub on the sides of this slit. Since this involves serious 
inconveniences, the film must be separated slightly from the slit 
as indicated in the figure. With the other illuminating apparatus 
the difference between the film and the exact position of the explor- 
ing zone may be due to an error in focusing. 

Since this difference in a general manner is different from zero 
and equal to d, a point M of the film near the axis receives light 
from all points of HH' provided that it is inside the cone HPH' 
whose peak P is obtained by joining the edges of the pupil and of 
the exploring zone located on the same side of the axis. The fully 
illuminated field on the film, therefore, has the height BB ' . But 
the film also receives a degraded illumination on the two bands, 
BC and B'C', the points C and C' being those where the film is 
struck by the rays which join the edges of the pupil and those of the 
exploring zone located on both sides of the axis. The band CC' 
is the total field which we shall call the explored zone. 

In order to calculate the luminous flux received by the point M 
of the explored zone (Fig. 6), the exploring zone is projected from 
point M on the plane of the pupil HH' ', and the common surface 
of the latter and the projection of the exploring zone is used. If 
the flux received by a point of the fully illuminated field is used as 
a unit, the flux received by the point M of the degraded field will 
be equal to the ratio between the area of the circular segment NiN^Hi 
and that of the circle 



10 LOUIS DUNOYER [J. S. M. P. E. 

Let us assume that the film gradually is moved away from the 
exploring zone; the fully illuminated field BB' is reduced to zero when 
the film passes through point P. For this position the illumi- 
nation decreases constantly from the center to the edge of the ex- 
plored zone CC' . When the film is above point P (Fig. 7), the parts 
taken by HH' and EE 1 in limiting the rays are exchanged, the ex- 
ploring zone now serving as outlet pupil and the surface HH' as a 
window. But the window surface employed decreases constantly 
when the point considered on the film moves away from the axis 
because, since the flux received by point P is used as a unit, the flux 
received by point M will be equal to the ratio of the circular seg- 
ment NiNzNi'Ni to that of the circle HyH^ and the area of this 
segment is maximum when NiNz and Ni'N* are symmetrical in 
regard to the center. The illumination of the explored zone de- 
creases constantly from the center to the edge. 

Let us calculate the heights h and h f of the field of uniform illumi- 
nation and the explored zone. Let D be the height of the outlet 
pupil HH', L its distance to the exploring zone, e the height of the 
latter, and d its distance to the film; finally d is the distance QP 
which is important, as we shall see. Triangles which evidently are 
similar give (Fig. 6): 



or since e always is very small compared with D: 

do = e (2) 

Then 

h d - d .D 

- = 5 or h = e d-f 
e do L 

or: 



and: 

h' d + QF D = L - QF 
e ~~ QF = QF 

from which: 



or: 



* - (' + Q 



Jan., 1932] LIGHTING OF SOUND FILMS 11 

If we should consider Fig. 7 instead of Fig. 6, we would find the 
same expressions for do and for h'; since there is no field of uniform 
illumination, h cannot be considered. 

It is useful to consider the values which the factor 

may assume. We call attention to the fact that since the flux which 
flows through the exploring zone should be as intense as possible, 

D 
the ratio should be large. If u is the opening half -angle of the 

beams which illuminate the exploring slit, we have: 

D 



The numerical aperture sin u of the illuminating apparatus in 
general should be at least 0.20 (u = 12 degrees) and with a micro- 
scope objective or a special photographic objective it may attain 

L 

0.4 (u = 24 degrees) which in the former case corresponds to 

L L 

= 2.5 and in the latter case to = 1.12; thus will be comprised 

approximately between 2 and 1. Consequently, according to for- 
mula (2) do will be comprised between twice the height of the ex- 
ploring zone and this height itself. Since this height may vary from 
0.02 millimeter to 0.05 millimeter, it is clear that d always will be 
very small, of the order of 0.02 to 0.1 millimeter. To assume that 
d may reach 10 times d Q , therefore, is not an inadmissible hypothesis. 
Moreover, we shall consider the maximum value. In any case an 
error in focusing which is entirely possible, amounting to 0.1 milli- 

d 

meter, for instance, may correspond to a value of of several units ; 

d 

and formula (4) shows that the explored zone can easily be several 
times greater than the exploring zone. 

5. Analysis of the Effects Produced by Enlarging the Explored 
Zone on the Film and by the Distribution of the Light in This Zone. 
We shall divide this analysis into two parts. We shall first ex- 
amine the effect of enlarging the exploring zone, of uniform illumi- 
nation, assuming that this zone is formed exactly on the film. We 
examine the influence of the width of the slit either when the film 
is in contact with the sides of it or when a perfectly focused image 
of a uniformly illuminated slit is formed on the film. This investi- 



12 



LOUIS DUNOYER 



[J. S. M. P. E. 



gation also applies to the case when apparatus without slit is used, 
provided that the illumination of the exploring zone is uniform. 
We also suppose that the exploring zone and the film do not coin- 
cide, and we shall examine the effect of the degraded illumination 
of the explored zone. 

In both cases it is necessary to form a hypothesis of the manner 
in which the transparency of the film varies along its length or, which 
is the same, the manner in which the total luminous flux would 
vary, which would go through the film if it were explored by means 
of a zone infinitely narrow in regard to the length occupied by a 
period of the transparency. In the case of a variable density film 
this transparency is the same as that of the film. For a constant 



minimum 



film 




\trans parency 
.. Y 



overage 

transparency 



FIG. 8. 



Curve representing sinusoidal variation of the film 
transparency along its length. 



density film it is defined by the ratio between the width of the part 
which is entirely clear and that of the opaque part. The most 
natural hypothesis which we can choose for the law of variation 
of the transparency of the film is that of a sinusoidal variation which 
would correspond to the perfect recording of a musical sound. In 
Fig. 8 the curve representing the variations of the transparency of 
the film along its length is plotted on the right. The abscissa XQ 
of the beginning of a period in regard to the axis of the lighting 
apparatus defines a given position of the film. Its transparency 
y at a point M, with the abscissa x, of the explored zone then will 
be expressed by: 

- x) 



y = a + b sin 



(5) 



Jan., 1932] 



LIGHTING OF SOUND FILMS 



13 



where / as in formula (1) is the length of a period of the transparency 
on the film, that is, the length occupied by the recording of a com- 
plete sound vibration. 

The constant b represents the half-amplitude of the variation 
of the opacity. It defines the amplitude or the intensity of the 
recorded sound. The constant a is equal to the mean value of the 
opacity when the film is unwound; in the photoelectric cell it corre- 
sponds to a constant current, hence is of no interest in regard to the 
sound. 

6. Effect of the Height of the Exploring Zone when Formed on 
the Film. Sound Efficiency. If the luminous flux which falls on the 




11 10 9 



7 6 54 



7 8 9 10 11 



3' 2 ^ J 2 3 if 5 

'Slit 
FIG. 9. Graphs of formulas (9), (10), and (11). 

film is used as a unit, the flux which leaves the film through a band of 
the height dx centered in M in the exploring zone will be expressed by: 

d& = ( a + b sin . J dx 

The total flux which leaves the film thus will be expressed by: 



or: 



_, . 

<b(xo) 



. bl . ire . 2-n-Xo 
ae H -- sm y sm j- 



(6) 



14 



LOUIS DUNOYER 



[J. S. M. P. E. 



Formula (6) shows that the leaving flux $(XQ) is a sinusoidal 
function of the abscissa XQ, of the same period / and the same phase 
as the transparency of the film. Whether the exploring zone is 
narrow or wide, the height of the sound is not changed and no har- 
monics will appear. 

In principle, therefore, it is not absolutely necessary to employ 
an extremely fine exploring zone to plot a band on which musical 
sounds are recorded, even very shrill ones, provided that they are 
sinusoidal. But, as we shall see, the intensity of the sound produced 
decreases rapidly when the width of the exploring zone increases 
and approximates the period of the transparency on the film. In 
other words, the sound performance of the latter decreases very 
rapidly. If a non-sinusoidal sound is concerned, having high har- 




II 



FIG. 10. Diagram for evaluating total flux passing 
through the film in a given position, XQ. 



monies, the timbre will be deformed, the more so according as the 
width of the exploring zone increases, because the harmonics will 
be the more reduced the higher they are. 

In reality, according to formula (6) the amplitude A of the varia- 
tion of the emerging flux will be expressed by: 

2-bl . TT6 

A = sin -7- 

TT I 



or 



. 7TC 

2bf sin 7- 

TTf I 



(7) 



If we assume that the ratio, -, of the width of the exploring zone 
to the period of the transparency decreases toward zero, it is clear 



Jan., 1932] LIGHTING OF SOUND FlLMS 15 

that the variations of the flux increase toward a maximum A corre- 
sponding to an ideally fine exploration of the band; we have 



The sound efficiency m of the exploring apparatus of the film can 
be defined by the expression: 

A I . TT6 

m = -r- = sm-r (8) 

A o TTC I 

d 
The curve = of Fig. 1 1 represents the sound efficiency plotted 

^ 

as ordinates as a function of the ratio - plotted as abscissae. It is 

If 

clear that the efficiency still is 90 per cent when the width of the 
exploring zone is equal to one-quarter period. It is only 30 per cent 
when the width of the exploring zone is three-quarters of a period. 
In order to calculate the frequency which is reproduced with a 
given efficiency, it is sufficient to eliminate N from formula (1) 

in which / is substituted by the value which gives to - the value indi- 

cated for the efficiency by the curve in Fig. 11. Thus, with an ex- 
ploring zone of 0.02 mm. the frequency obtained with an efficiency 

/e 1 \ 

of 90 per cent will be ( - = - , hence / = 4-0.02 ): 



a frequency which is slightly higher than that of re-i of a piccolo. 
The efficiency corresponding to this note will be exactly 94 per cent 



o - 



Another example: the frequency corresponding to an efficiency 
of 25 per cent will be: 

455- 0.785 = 
0.02 

e 0.02 \ 

from which: - = 0.785 and consequently / = ;r^J- 
/ 0.785/ 

The frequency 20,000 ( *- = : = 0.88 J will be reproduced 

only with an efficiency of 13 per cent. 



16 



LOUIS DUNOYER 



. S. M. P. E. 



We note that probably a much higher sound quality could be 
obtained in regard to the reproduction of the voice and of timbres 
if the width of the exploring zone dropped to 0.01 millimeter. The 
frequency 20,000 then would be reproduced with an efficiency of 
70 per cent. This is due to the very rapid decline of the efficiency 




0,1 



FIG. 11. Curves showing the sound efficiency m as a function of K = c/l for 

various values of d/d . 

indicated by the form of the curve when the width of the exploring 
zone exceeds one-quarter of the period of the transparency on the 
film. 

7. Effect of an Error in Focusing or of a Reduced Explored Zone. 
We refer to Figs. 6 and 7 for the investigation of this problem, The 
calculation of the luminous flux leaving the film, for a given posi- 



Jan., 1932] LIGHTING OF SOUND FlLMS 17 

tion of the latter, presupposes that the distribution of the incident 
flux is known. In order to calculate the flux which falls in a point 
M of the abscissa x in regard to the axis of the illuminating appara- 
tus, we project, as mentioned above, the exploring zone on the plane 
of the outlet pupil HH' of the illuminating apparatus, and we con- 
sider the common surface of this projection and this pupil. Let S 
be this surface, which has been shaded in Figs. 6 and 7. Since the 
flux illuminating the fully illuminated zone is used as a unit, the flux 
illuminating the point M will be equal to the ratio between this 
common surface S and the surface wR 2 of the pupil: 



Calculation by integration of the surface S leads to the following 
results: 

In the case of Fig. 6 (d < do) we have in the fully illuminated 
field: 



and in the reduced part of the explored zone, that is, for: 

1 <_<<*<!+< 
o <*o 

5 1 , . d, /, 2x 

_ = + arcsm 



In the case of Fig. 7 (d > do) we obtain two different expressions 
for 5 according as the projection of the exploring zone goes through 
only the outlet pupil or projects beyond it. In the former case 
we obtain: 



1 . d /, 2x\ . 1 . d A , 2x , \ 

- arc sin -^ ( 1 ) + - arc sin -f [ 1 H h I 

x d \ e/7T d \ e/ 



S 
--arc sn 



with: 

2* 



18 LOUIS DUNOYER [J. S. M. P. E. 

and in the latter case: 



with: 



*_!<*<*+, 

d d 



The formulas (9), (10), and (11) essentially assume that x is posi- 
tive. It is clear that the incident flux $i(x) is the same for two 
symmetrical points in regard to the axis. For negative x, therefore, 
x should be replaced by x in the second terms of the formulas 
(9), (10), and (11). 

These formulas have been translated into curves (Fig. 9) for dif- 

d 

ferent values of . 
do 

On the other hand, the transparency of the film as above is as- 
sumed to be expressed by the second term of formula (5) at point 
M of the film (Fig. 10). The luminous flux passing through the 
film in M through a band of the height dx is: 

$C 



The total flux flowing through the film which is placed in a given 
position, that is, for a given value of XQ, will be obtained by inte- 
grating this expression from one limit to the other of the total field 
or explored zone which has the height h'. (See 4.) We have, there- 
fore: 



+ - 
= f J *<(*) (a 



b sin dx 



Like the transparency of the film this function of x has the period /. 
This is evident physically; this also is due to the fact that if x in- 
creases from /, the function under the sign of integration does not 
change and that both limits of integration increase from /. The 

emerging flux $,(#0) also goes through a maximum for x = - and 

31 4 

through a minimum for XQ = ; we have in reality: 

4 



Jan., 1932] LIGHTING OF SOUND FlLMS 19 



.*' 

/ 2" 2-n-b 
V T*** 



For #o = ~ we have: 

4 



, 

. 27T* 

m 



and since $j(#) assumes equal values for equal values of x and oppo- 
site signs, it is clear that the preceding integral is zero. The same 

31 

holds for #o = T The emerging flux & e (x Q ), therefore, has maxima 
4 

and minima in the moments when the axis of the illuminating ap- 
paratus goes through the film in a region of maximum or minimum 
transparency. Finally, this flux oscillates around the same mean 
value no matter what the extent of the explored zone may be. If, 
therefore, we put XQ = 0, we have: 

,h r h' 

C 2" / 2-*X\ C " 

/ ^ <*><(*) ( a - b sin -- J dx = a j / 
~ ~2 "2" 

The last integral represents the total flux illuminating the ex- 
plored zone which is assumed to be constant and equal to 1 for any 
surface of this zone. 

Since the curves represent the variations of the flux which has 
passed through the film, as a function of XQ, that is, of the position 

d 

of the film and for every value of , they are undulating curves of 

d Q 

the same period and the same phase as the transparency of the film 
(with one exception which will be examined below), all having the 
same average ordinate. These curves naturally are a function of 
the coefficients a and b, but the value of the efficiency of the illumi- 
nating apparatus defined as above is not. In reality the amplitude 
A of the variations of the flux, according to what has been said, is 

equal to the difference between the values of $ e (x ) for XQ = - and 

#o = Hence: 
4 



20 LOUIS DUNOYER [J. S. M. P. E. 

/?' 
*(x)co8^dx (13) 

2 





= 2b I 2 4>.-( 
J- h - 



If the explored zone having the height h f were infinitely narrow 
in regard to the period / (perfect exploration of the film), the ampli- 
tude of the variations of the flux would be as we have seen in the 
preceding paragraph: 

A, = 2be (14) 



The efficiency w, therefore, will be expressed by: 

i c +h * 2 

m = - I fy(x) cos j- dx (15) 

^-2- 
With 



If we consider the formulas (9), (10), and (11), we find 

2x 
really is a function of . If we write, therefore, 



^ = X K = - (16) 

we get: 

m = I 3>i(xY cos KirX -dX (17) 



a formula which is independent of the coefficients a and Z> which 
enter into the law of variation of the transparency. 

In order to calculate the efficiency according to formula (17), 

d 

a value of K first must be chosen. Choosing a given value of 

do 

we calculate for a sufficient number of values of X the ordinates of 

the curve $i(X) corresponding to this value of ; we use the formulas 

do 

(9), (10), and (11), bearing in mind that when there is a fully illumi- 
nated field $i(X) = 1 in this field. Thus the curves of Fig. 10 are 
plotted. The ordinates of these curves are multiplied by cos KirX 
and the curve fy(X) cos KtrX is plotted. Then only the area which 
it defines above the axis of X remains to be measured. 



Jan., 1932] 



LIGHTING OF SOUND FILMS 



21 



The curves shown in Fig. 1 1 which represent the efficiency m as a 
function of K have been constructed in this way, each one being 

plotted for a given value of . The uppermost curve corresponds 

d 

to a perfect focusing (d = 0). We have already examined it in the 
preceding paragraph. 



0.9 



0,8 



0,7 




0.5 




0.3 



.0,2 



0.1 - 





10 



FIG. 12. Curves derived from those of Fig. 11, showing the variation of the 
efficiency with the focusing, i. e., with d/d the ratio K remaining constant. 

The curves shown in Fig. 12, which are derived from those in Fig. 
11, indicate how the efficiency varies when the focusing is varied, 

that is, , the ratio K remaining constant as for a given illuminating 
d 

apparatus and a given film. 



22 Louis DUNOYER [J. S. M. P. E. 

It appears from the curves in Figs. 11 and 12 that the poorer the 

/ d\ 

focusing is ( large value of J, the more rapidly the efficiency de- 

V do/ 

creases when the ratio between the width of the exploring zone and 
the period of the transparency decreases. 

The analysis of these curves clearly demonstrates the great im- 
portance of the focusing. Let us assume, for instance, that the 
note recorded on the film is 7^7 (4698 vibrations) and that the width 

of the exploring zone is a quarter period, K = - (I 0.097 mm., 

4 

e = 0.024 mm.). We have seen already that for perfect focusing 
the sound performance will be 90 per cent. If we assume that the 
aperture of the illuminating pencil is 60 degrees (D = L), we have 
do = e = 0.024 mm. An error in focusing of only 0.1 millimeter 

d 
will give = 4. The curve in Fig. 1 1 corresponding to this value 

d d 1 

of shows for the abscissa K = - that the sound efficiency drops 
do 

to 17 per cent, that is, less than one-fifth of the value which it had 
with perfect focusing. 

8. Remarks on the Case when the Explored Zone Covers Several 
Periods of the Transparency. Each one of the curves in Figs. 11 
and 12 is limited to an arc comprised between the efficiency limit 
1 and the efficiency zero. They could be extended beyond this. 

According to formula (17) for a given value of , m starting from 1 

d 

for K = decreases when K increases and reaches the value 0. 

For = 4, for instance, we have m = for - = 0.31. If- increases 
do II 

still more, m becomes negative. This is not surprising when we 
consider the formulas (13), (14), and (15). Formula (13) particu- 
larly shows that if - or K exceeds the first value for which the ampli- 
tude A is zero, the sign of the latter changes, that is to say, the 
emerging luminous flux still has maxima and minima but in phase 
opposition with the transparency of the film at the point where it 
meets the optical axis. ' The efficiency then is. equal to the absolute 

value of m. When - continues to increase, this absolute value goes 



Jan., 1932] 



LIGHTING OF SOUND FILMS 



23 



through a maximum, returns to zero, then again assumes positive 
values and so on. 

It is clear that the successive maxima always are decreasing. 
Let us assume that d = (exploring zone on the film), for example. 

The efficiency m decreases from 1 to zero when - increases from to 1. 

If the exploring zone is further enlarged, the emerging flux begins 
to fluctuate again; their amplitude will be maximum when the 
exploring zone covers P/2 period on the film; the emerging flux 
will be maximum when the optical axis goes through the film in a 
minimum of transparency as shown in Fig. 13 (a) ; it will be minimum 



Optical axis 




Optical axis 




FIG. 13. (a) Showing how the emerging flux is a maximum when the 
optical axis goes through the film in a minimum of transparency, and 
(6) how it is a minimum when the axis goes through a maximum of trans- 
parency. 



when the axis goes through a maximum of transparency (Fig. 
The fluctuation again will be zero when the exploring zone covers 
two periods of transparency; then if the exploring zone is further 
enlarged, other fluctuations will result with a maximum when it 
covers an odd number of half-periods and again become zero when 
it covers a whole number of periods. It is also clear that the dif- 
ference between the maximum and the minimum of the flux, that 
is, the amplitude of the fluctuations, will decrease when the num- 
ber of periods of the transparency simultaneously involved increases. 
Consequently an adjustment of the focusing and the width of the 
exploring zone, which produces an efficiency equal to zero for a given 



24 LOUIS DUNOYER [J. S. M. P. E. 

frequency, will produce an efficiency which differs from zero for 
higher frequencies. In the case of perfect focusing, for instance, 
the efficiency according to formula (8) will have maxima for: 

c = 3 5 7 2n + 1 

I ~ 2' 2' 2 2 ' 

and the values of the efficiency will be, respectively: 

I - ' 21 - I - ' 13 ' T, ~ ' 09 ' I - ' 07 

The corresponding frequencies will be given by formula (1) when 
e has been chosen. An exploring zone of 0.1 millimeter would give, 
for instance: for N = 4550, 9100, 13,650, etc., an efficiency = 0; 
and for N = 6830, 11,390, 15,900, 20,450, etc., efficiencies equal to 
21, 13, 9, 7 per cent, etc. 

This example clearly shows the disturbance or unbalance which 
a slightly wide exploring zone can produce in a symphonic reproduc- 
tion even with perfect focusing. Such disturbances are inadmissible. 
For this reason we have in paragraphs 6 and 7 systematically limited 
the investigation of the efficiency to the range between its maxi- 
mum limit and its first zero minimum. The remark, which we just 
have made, should be borne in mind and it no doubt could explain 
certain sound distortions actually observed. 

9. Effect of a Lack of Uniformity in the Illumination of the Ex- 
ploring Zone. Heretofore we have assumed that the luminous 
flux flowing through the exploring zone was the same at every point. 
With some methods of illumination this is not so: for example, if 
the image of a spiral incandescent filament with the turns spaced 
too far apart is formed in the plane of the exploring zone, or if an 
error in centering the optical parts causes a lack of symmetry in the 
illumination of this zone, or if this zone is the image of a slit whose 
edges are not parallel or if this slit is partially closed. 

This lack of uniformity presents the greatest inconveniencies 
for constant density films. If we assume, to consider the extreme 
case, that only a part of the width of the sound band (Fig. 14) is 
swept by the exploring zone, it is clear that only the peaks corre- 
sponding to the most intense vibrations produce fluctuations in the 
flux transmitted through the film and that the resulting sound will 
only be remotely related to the recorded sound. Without going to 
this extreme, it is clear that any lack of uniformity in the illumi- 



Jan., 1932] 



LIGHTING OF SOUND FILMS 



25 



Sound band 



nation of the exploring zone will favor certain parts of the recording 
curve and consequently certain sounds at the expense of others. 
A more or less strong sound distortion will take place. 

On the other hand, this lack of uniformity has no importance for 
variable density films since the transparency of the film is the same 
on its entire width. Then only a difference in height between the 
ends of the exploring zone is detrimental (edges of slit not parallel), 
but less, of course, than if this zone had the height of the widest end 
on the entire width of the band. 

For an equal height of the exploring zone the variable density 
films, therefore, are much less sensitive to the im- 
perfections of the illuminating apparatus than 
the constant density films. 

10. Resume of Part I. The essential points of 
the analysis which we have outlined above are 
summed up in the formulas (1), (2), (3), (4), (8), 
(9), (10), (11), (16), and (17). We have intro- 
duced the important idea of the sound efficiency 
of an illuminating device connected to a film 
which is supposed to be perfect, in the same way 
as all the devices which actually transform the 
fluctuations of the luminous flux passing through 
the film into sound vibrations are supposed to be 
perfect. The variations of the sound efficiency 
as a function of (1) the ratio between the width 
of the exploring luminous zone and the period of 
the transparency on the film and (2) the focusing, 
are represented by the curves in Figs. 11 and 12 
which allow of determining primarily the efficiency 
under any given practical circumstance. Not only 
do they admit of calculating this efficiency for a pure sound (sinusoidal) 
of a given period but in the case of a fundamental sound accompanied 
by various harmonics they also admit of calculating the ratios in 
which these diverse composing vibrations will be reproduced. Thus 
they completely solve the problem of the sound distortion produced 
by a given illuminating apparatus. 

These curves particularly demonstrate how rapidly the efficiency 
decreases as a result of an error in focusing or the widening of the 
exploring zone. We shall apply the results of Part I of our paper 
to the investigation of different types of lighting apparatus. 




Exploring 
zone 

FIG. 14. Illustrat- 
ing the case where 
only part of the 
sound band is swept 
by the exploring 
zone, due to un- 
uniform illumina- 
tion of the zone. 



26 Louis DUNOYER [J. S. M. P. E. 

PART H 
DIFFERENT LIGHTING APPARATUS 

In section 3 of Part I we enumerated succinctly the different 
types of lighting apparatus in order to .base the theoretical investi- 
gation of the lighting of sound films on sufficient concrete data. 
In Part II we shall not describe them in detail but study their opera- 
tion. 

11. Apparatus with Slit near the Film. The simplest method 
of powerfully illuminating a very narrow zone of the sound band 
evidently is to place the film on the sides of a very fine slit and illumi- 
nate the latter strongly. To produce this illumination a source 
could be provided, so extensive, or placed so near the slit, that the 
angle under which the center of the latter is seen would be very 
great. If the source itself is very brilliant, it is clear that a more 
intense illumination is obtained in this manner than with any opti- 
cal system. The apparatus at the same time would be extremely 
simple. 

The available sources of great brilliancy, however, have a high 
temperature and cannot be placed sufficiently near the film. A 
condenser (Fig. 4) must be used which concentrates the light on the 
slit in forming a more or less good image of the source on the slit. 
If this image were perfect and the condenser did not absorb light, 
each of the surface elements of the image would have a brilliancy 
equal to that of the conjugate surface element of the source. In 
reality the reflection on the glasses of the condenser, the absorption 
of light by the latter, and the aberrations of this apparatus, which 
in general are considerable, materially reduce the effective brilliancy 
of the image formed on the slit. On the other hand, when the source 
is a spiral filament, as usually is the case, the aberrations have the 
effect of making the illumination of the slit uniform, a fact which 
is valuable for a constant density film. (See 9.) 

The very serious disadvantage of these devices is that if the film 
rubs even very slightly on the edges of the slit, it is rapidly scratched 
and, besides, the slit soon is closed by dust. This dust cannot be 
avoided even if the sides of the slit are polished mirror-like and 
curved inward so as to touch the film only at points separated 
somewhat more than the width of the slit. Irregularly accumulated, 
it also can be carried along suddenly; thus it causes sudden varia- 
tions of the luminous flux illuminating the film and hence inad- 
missible interfering noises. 



Jan., 1932] LIGHTING OF SOUND FILMS 27 

In order to avoid them, a small space could be left between the 
slit and the film. A considerable decrease of the sound efficiency 
would result, however. In reality, we noted in 4 that the critical 
distance d G is the of same order as the width of the slit if the angular 
opening of the pencil which illuminates it is slightly large, as is 
necessary (24 degrees at least). If quite high frequencies are to be 
explored when the film is illuminated with sufficient intensity, d Q 
never will be far from 0.02 to 0.05 millimeter. If the film does not 
touch the sides of the slit, the air current which it produces near 
its surface and its electrification carry along atmospheric dust parti- 
cles which also adhere to the edges of the slit unless the space be- 
tween the latter and the film is sufficiently large. We consider a 

d 

space d of 0.1 millimeter as a minimum. The ratio , therefore, 

d d 

will be comprised between 5 and 2. If = 4, for instance, the ef- 

d Q 

ficiency, which is equal to 80 per cent when the height of the slit is 
0.1 per cent of the period of the transparency of the film, drops to 
17 per cent when the height of the slit reaches one-quarter of a period, 

and to when - = 0.32. That is to say, with a slit of 0.025 milli- 
meter, a space d = 0.1 millimeter, an angular opening of the illumi- 
nating pencil of 45 degrees (d Q = 0.025 mm.), the efficiency is 80 
per cent for a frequency of 1820 (approximately la& sharp), drops 
to 17 per cent for a frequency of 4550 (approximately sharp do^ 
and to for a frequency of 5820. 

The dust could be avoided entirely by leaving only a space of less 
than 0.1 millimeter between the film and the slit but this makes the 
slit somewhat complicated. It could be covered with a film. In 
order that neither be scratched no other material than glass should 
be chosen. Thus a thin glass plate is attached to the sides of the 
slit by means of a suitable adhesive (Canada balsam, for instance); 
then the external face of this plate is ground by processes ordinarily 
used by opticians in order to reduce its thickness as much as possible. 
We do not know whether the system has been employed effectively 
for illuminating films during reproduction, but an entirely similar 
device is used particularly by the Fox Movietone for the sound re- 
cording of the film. The distance between the slit and the film 
thus can be reduced to a few microns. 

12. Apparatus with Projected Slit, In this type of apparatus 



28 



LOUIS DUNOYER 



[J. S. M. P. E. 



an image of the slit is formed on the film by means of an objective 
(in general, a microscope objective). Since this image is smaller 
than the slit, the latter may be wider. Being separated from the 
film by the objective it can be placed in a closed space. For these 
two reasons dust is much less to be feared. The slit is illuminated 
by means of a lamp with a spiral filament, of the automobile head- 



FIG. 15. A reproduction, enlarged approximately 14 times, of the 
image formed on the film in a high-grade apparatus with a slit which 
had been used only a short time. Note the breaks in the image caused 
by dust in the slit. 

light type, with a condenser interposed. Let us analyze more closely 
the conditions of construction which obtain for this type of apparatus. 
First, the necessary ratio between the length of the image (or 
exploring zone) and its width requires a very fine slit. It must 
not be forgotten that the width of the sound band is three milli- 
meters and that the height (or width) of the exploring 
be approximately 0.02 to 0.05 millimeter, the latter dimension, 
moreover, being too large and acceptable only as a makeshift. Hence 




FIG. 16. One means of lighting the slit consists in 
forming an image of the filament on the slit by means of a 
condenser. 

the slit itself should be from 100 to 150 times longer than it is wide. 
A slit of 0.5 millimeter should be from 50 to 75 millimeters long, 
which is difficult to employ owing to the space required and the 
difficulty of illuminating it sufficiently. In reality we have to use 
slits which are 0.1 millimeter wide and consequently 10 to 15 milli- 
meters long. 



Jan., 1932] 



LIGHTING OF SOUND FILMS 



29 



Experiments show that dust from the air very easily clings to the 
edges of a slit 0.1 millimeter wide. Fig. 15 is the reproduction, 
enlarged approximately 14 times, of the image formed on the film 
in a high-grade apparatus with slit 
which has been in use only a little 
as yet. We note, however, that 
this image is cut six times by dust 
which has fallen on the slit. The 
latter was 12.5 millimeters long and 
0.1 millimeter wide; a microscope 
objective formed an image of it 4.6 
times smaller. The exploring zone thus was 0.022 millimeter wide. 
In order to obtain the photograph reproduced in Fig. 15 this image is 
retaken by means of a photographic objective. 

In Fig. 15 we also find that the illumination of the image is little 
uniform from one end to the other. The centering had been par- 
ticularly careful, however. The slightest lack of adjustment in- 
creases this lack of symmetry materially, which depends on the 
manner in which the slit is illuminated. 

Among the methods of lighting the slit, two should be particu- 



FIG. 17. Photograph of a coiled 
filament lamp, showing variations of 
brightness between turns. 




FIG. 18. Another method of illuminating the slit 
consists in forming the image of the filament on the inlet 
pupil of the microscope objective. 

larly considered. One consists in forming an image of the incan- 
descent filament on the slit by means of the condenser (Fig. 16). 
The other (Fig. 18) consists in forming the image of the filament 
on the inlet pupil of the microscope objective; in the very same 
manner as when the image of a slide is to be projected on a screen, 
the image of the source of light is formed on the projection objec- 
tive by placing the slide very close to the condenser. 

In the former case the illuminated part of the film is the image 
of the part of the conjugate source of light of the slit in regard to 
the condenser. The illumination of the exploring zone, therefore, 



30 LOUIS DUNOYER [J. S. M. P. E. 

is uniform only if the brilliancy of the source is uniform within this 
part of the source. This is not so when it consists of a helical fila- 
ment even when the turns are so close that the images of the back 
half-turns are formed between those of the front half -turns because 
the temperature of the internal surfaces of the turns is higher than the 
temperature of their external surfaces. This is illustrated in Fig. 
17 which reproduces the photograph of the filament of the lamp 
(automobile headlight type) used in the apparatus with slit to which 
Fig. 15 refers also. 

On the other hand, in the case when the image of the source is 
formed through the slit on the microscope objective, the illumi- 
nation of the exploring zone, image of the slit, is perfectly uniform. 
In reality, each point M of the exploring zone is illuminated by the 
continuous flux in a cone whose peak is the conjugate point M' 
of M on the slit and the base on the objective is the part of the 
image of the filament limited by this objective (or more accurately, 
by its inlet pupil). 

Evidently the former method of illumination is inadequate for 
illuminating a constant density film since for a film of this type the 
illumination of the exploring zone should be uniform (9). This 
condition is not important for a variable density film. Neverthe- 
less, the latter method of adjustment in general is used even for 
these films. In order that the total flux which falls on the film 
then can be as great as with the former method of adjustment, the 
microscope (or more generally projection) objective must be covered 
entirely by the image of the filament which the condenser produces. 

In order to calculate it we shall neglect the losses of light due to 
absorption, reflection, and diffusion through the lenses and indicate 
the brilliancy of the source by B. In the first mode of adjustment 
(Fig. 16) the flux falling on the film is expressed by: 

Gi^B-hS (18) 

where kiS is the surface of the part of the image of the source within 
the slit, with surface S, S the surface of the entrance pupil of the 
objective 0, and q the distance from the slit to this objective. In 
the second mode of adjustment (Fig. 18) the expression of the flux 
falling on the film is: 



Jan., 1932] LIGHTING OF SOUND FlLMS 31 

when k z S is the part of the surface of the objective covered by the 
image of the source which is formed on it by the condenser. The 
problem then is to know whether k\ is larger or smaller than k 2 . If 
the turns of the filament are sufficiently close so that the image of 
the back half-turns is projected between the images of the front 
half- turns, the slit is covered completely and we have ki = 1. The 
flux $2, therefore, cannot be greater than the flux $1 but it can be 
equal to it if k z 1, that is to say, if the entrance pupil of the pro- 
jection objective is completely covered by the image of the filament 
formed by the condenser. 

Nevertheless, much light is lost with both methods. In the 
former method all the light is lost which forms the part of the image 
of the filament outside the slit and all the light which, after having 
passed through the slit, does not reach the objective 0, the angular 
aperture of the condenser in general being larger than that of the 
projection objective. In the latter method of adjustment all the 
light is lost which does not go through the slit and, besides, all 
the light which, having passed through the slit, will pass through the 
points of the image of the filament located outside the entrance 
pupil of the projection objective. 

From the above it results that the devices with projected slit 
always utilize only a small portion of the light projected by the 
source on the condenser and that their efficiency is low. Numerical 
data bearing on this subject will be given later on. 

13. Apparatus without Slit, with Cylindrical or Cylindrospherical 
Lenses. In order to avoid the disadvantage of dust on the slit, a 
disadvantage which the devices with projected slit do not avoid 
completely as we have just observed, and the losses of light involved 
in these devices, two solutions have been proposed. The first one 
is the solution which we have adopted and with which we shall con- 
clude this article. But, first, we shall describe briefly the other solu- 
tion which is based on the use of cylindrical or cylindrospherical 
lenses. 

It is well known that the luminous rays sent from a point located 
on the axis of such a lens, after having passed through it, strike two 
perpendicular focal lines, one of which is parallel to the generators 
of the cylinder. A spiral filament whose axis is parallel to the genera- 
tors also concentrates the light on two luminous bands, of which 
the one which is nearest the lens, and consequently is the narrowest, 
also is parallel to the generators. This small luminous band may 



32 Louis DUNOYER [J. S. M. P. E. 

replace the slit, and a projection objective (in general, a microscope 
objective) forms an image of it on the film. 

Since the rays issued from every point of the source outside the 
plane of the principal section which is perpendicular to the genera- 
tors and the plane of the principal section parallel to the latter pass 
through different points of the focal distance, the distribution of the 
flux in the small luminous band furnished by the cylindrospherical 
condenser is independent of the distribution of the intensity in the 
source and very uniform on the entire useful length of this band. 
This fact as well as the elimination of dust makes the device very 
attractive for constant density films. 

Its disadvantage seems to be the difficulty of obtaining a sufficiently 
fine exploring zone for reproducing high frequencies with a sufficiently 
intense useful flux. If the small luminous band produced by the 
cylindrospherical condenser at the slit of the apparatus with the pro- 
jected slit examined above is substituted, we note that this band 
should be 12.5 millimeters long and only 0.1 millimeter wide on the 
entire length. It seems difficult to obtain such an image with a 
spiral filament, which always would have a diameter of 2 to 3 milli- 
meters at least, and with beams which should have a numerical 
aperture of at least 0.2. If it is possible to correct the aberrations 
in the plane of the principal section which is perpendicular to the 
generators in such a way that the image, furnished by the rays com- 
prised in this plane would be only 0.1 millimeter wide, it seems 
improbable that the rays which are not comprised in this plane 
and are oblique to the generators could be regulated with the same 
precision. 

14. Apparatus without Slit, with Rectilinear Filament Lamp. 
There is an extremely simple means of avoiding the dust, the losses 
of light on the sides of the slits, the aberrations of the cylindrical 
lenses, and the lack of uniformity of the brilliancy on the spiral 
filaments. It consists in constructing the lighting apparatus with 
a lamp which has only one filament set up rectilinearly and forming 
a reduced image of this filament on the film. If the filament is very 
fine and the aberrations of the optical system which forms its image 
are well corrected, it is clear that the exploring zone will be as fine 
as the separating power of the objective employed will permit. 
There is nothing to prevent it from being reduced to the width of 
the finest details which can be detected by a microscope, that is, 
less than 1 micron. According to 6 such a fine exploring zone 






Jan., 1932] LIGHTING OF SOUND FlLMS 

G 



33 




H K 



FIG. 19. Lighting apparatus with rectilinear filament for sound 
films. A, lamp; B, rectilinear filament; C, closing mirror soldered 
into the wall of the lamp; D, elastic plate used for guiding the 
lamp when introduced in the apparatus; E, objective; F, ventila- 
tion hole; G, window for letting out the luminous pencil, diminating 
interfering reflections; H, swivel joint; /, socket of swivel joint; 
/, lamp holder tube for orientating the filament; K, covers of 
spikes of current supply; L, tube forming the body of the appara- 
tus; M, tightening screw making the tube /immovable; N, screw 
for centering the filament; 0, focusing ring. 



34 



LOUIS DUNOYER 



[J. S. M. P. E. 




makes it possible to reproduce vibrations of 120,000 cycles with an 
efficiency of 90 per cent, that is, well above the audible range. More- 
over, since all the light falling on the entrance pupil of the projec- 
tion objective is used to form the image, except for losses in this 
objective, the optical efficiency of the apparatus will be excellent. 
Eliminating the condenser will reduce the losses of light still more. 
Figs. 19 and 20 show the first model of the apparatus which we 

had made according to this principle 
by the Societe S. C. A. D. and the 
lamp which it contains. 

The filament of this lamp, which is 
of tungsten 25 mm. long with a di- 
ameter of 0.1 mm., is stretched be- 
tween two metallic rods which con- 
duct the current. The difficulty of 
constructing this lamp lies in the 
choice of the metal constituting these 
rods and in the tension which the fila- 
ment should have. If it is stretched 
too much, it breaks at a high tem- 
perature ; if it is not stretched enough, 
it does not stay rectilinear when it is 
brought to incandescence. Methods 
of construction have been perfected 
so that the filament remains perfectly 
rectilinear at its normal operating 
temperature (2290K.) and at the 
same time its tension is low enough 
not to jeopardize its life, which is 
several hundred hours. 

In order to obtain a rectilinear 
image of the filament and a careful 

correction of the aberrations, the rays employed must go through 
the walls of the bulb under conditions which are known perfectly. 
For this purpose the bulb of the lamp consists of a glass cylinder 
closed at the end opposite the base of the lamp by a lens with parallel 
faces ground optically and of a known thickness. By means of a 
special method this lens is fused into the walls of the cylinder, the 
deformations resulting from the junction not extending to the central 
part through which the useful rays are passing. 



FIG. 20. Photograph of lamp 
used in apparatus shown in 
Fig. 19. 



Jan., 1932] LIGHTING OF SOUND FlLMS 35 

These rays fall upon a photographic objective of the anastig- 
matic triplet type. An objective of this type has been preferred 
to a microscope objective because the enlargement which is to be 
obtained is approximately */4 in order that the exploring zone shall 
be 3 mm. long and 0.0125 mm. wide, which is considered sufficiently 
small at the moment. This width in reality is half as great as the 
width of most apparatus with a slit and still gives a sound efficiency 
of nearly 70 per cent (see 6) for the frequency 20,000, whereas 
twice the width gives an efficiency of only 13 per cent for the same 
frequency. The microscope objectives corrected for this magnifica- 
tion and an object field exceeding 3 mm. in general have a numerical 
aperture which is lower than 0.15, whereas anastigmatic triplet lenses 
are found to have an excellent definition in a field exceeding by far 
3 mm. and having a numerical aperture of 0.25 at least. Moreover, 
with such objectives the distance separating the last lens from the 
film is greater than with ordinary microscope objectives, which also 
may be considered an advantage. 



FIG. 21. The image of the filament. 

Fig. 21 shows the image, on the same scale as the image of the 
slit in Fig. 15, of the image of the filament as received by the film. 
It can easily be proved that this image is perfectly rectilinear. The 
brilliancy of the exploring zone then will be perfectly uniform ex- 
cept in the immediate vicinity of the ends, whereas it is far from 
uniform with the apparatus with the slit, although adjusted with 
care, as shown in Fig. 15. 

Naturally it would not be difficult to reduce the width of the ex- 
ploring zone still more either by using a finer filament of the same 
length with the same objective and the same magnification or by re- 
ducing the magnification and increasing proportionally the length of 
the filament, which is assumed to have a constant diameter. 

In the model shown in Fig. 20, the socket of the lamp is provided 
with spikes which sink into an insulating piece provided with a 
swivel joint H whose case is in one piece with the plug K fixed on 
the tube / which can rotate with slight friction against the external 
tube L. By turning the plug K the filament can be made hori- 



36 Louis DUNOYER [J. s. M. P. E. 

zontal (the film being unwound vertically). Four screws N whose 
points rest on the walls of the bulb through springs D make it possible 
to center the filament in such a manner that the exploring zone 
sweeps the sound band exactly. The blades of the spring also 
facilitate the introduction of the lamp mounted on the plug K in 
the tube L. 

The pencil illuminating the exploring zone goes through the small 
window G without touching its edges, this window being used only 
to eliminate the interfering light reflected on the internal walls of 
the tube L. 

Focusing on the film is performed by means of the ring which 
displaced the objective or an element of the objective by small 
amounts. 

15. Results of Experiments. In addition to the satisfactory 
results obtained in operation with this apparatus we have tried 
to determine the total luminous flux which it sends through the 
exploring zone and its energy output, that is, the number of watts 
consumed to obtain this luminous flux. For comparison this in- 
vestigation was conducted also with the apparatus with slit, already 
mentioned. 

Either of these two lighting devices was fixed on a rotating sup- 
port, the axis being vertical, so that the entire emerging flux was 
received by a photoelectric cell (hemipherical S. C. A. D. cell), and 
by one rotation of the support this flux could be substituted instan- 
taneously by the flux from a standard lamp mounted on a rack 
support. The cell as usual was connected to a battery and a galva- 
nometer. The lamps, that of the investigated lighting apparatus 
as well as the standard lamp, were supplied by a storage battery 
with potentiometers to regulate with precision the current in the 
lamps, standard ammeters to measure it and standard voltmeters 
to measure the voltages at the terminals of the latter. By experi- 
menting, the distance of the standard lamp was regulated in such 
a manner that the deflection of the galvanometer was the same 
when the cell received the flux leaving the lighting apparatus or 
that from the standard lamp limited by a diaphragm of a known 
surface placed on the cell. The standard lamp was supplied in 
such a manner that its potential at the terminals was its standard 
potential, 102.9 volts; its luminous intensity then was 20.2 candles. 

The results of a measurement made on our apparatus and on the 
apparatus with the slit are as follows: 



Jan., 1932] LIGHTING OF SOUND FlLMS 37 

Apparatus Apparatus 

L. D. with Slit 

Current in the lamp, amperes 1.5 5.5 

Voltage at the terminals, volts 3.44 7.69 

Power consumed, watts 5.16 4.24 
Diameter of the diaphragm placed on the cell illuminated 

by the standard lamp, centimeters 2.535 2.043 
Distance of the standard lamp for the equilibration, 

centimeters 74.3 67.3 

20.2 x 2.S35 2 

Flux leaving the lighting apparatus 

4-74.S 2 

20 2 7T 2 043 2 

(L.D.); - (with slit); lumen 0.0185 0.0146 

4-67.3 2 

Before we conclude this paper we wish to make the following re- 
marks in regard to these results: 

(1) Variations in the centering of the lamp of our apparatus pro- 
duce no effect on the emerging flux, provided, of course, that the 
decentering is not so great that the pencil is partly hidden by the 
edges of the window G. On the other hand, very slight variations 
in the centering of the apparatus with the slit produce very great 
variations of the emerging flux. This is readily understood since the 
region of the spiral filament is varied, the image of which is formed 
by the condenser on the microscope objective. The flux obtained 
above is the maximum flux which we have been able to produce; 
a very slight irregularity which leaves an excellent centering upon 
examining the pencil makes the flux drop to 0.0122 lumen. 

(2) If the energy output is expressed by the number of lumens 
emitted in the emerging pencil for 1 watt consumed, it is clear that 
the apparatus with rectilinear filament, which already is superior 
to the apparatus with slit in absolute value, is far superior in regard 
to efficiency. Its efficiency is 0.0036 lumen per watt, whereas the 
efficiency of the apparatus with slit is one-tenth as great, that is, 
0.00034 lumen per watt. Such a result could be expected owing 
to the losses of light avoided in the apparatus with rectilinear fila- 
ment. This advantage, which perhaps is little appreciated in the 
present talking picture installations, evidently may attain great 
importance. 

(3) The efficiency of the apparatus with rectilinear filament would 
be increased still further in regard to that of the apparatus with 
slit if our lamp were as powerful as that of this apparatus. With 
1.5 amperes the color temperature in the middle of the filament 
of our lamp is 2290 K., the output then being 7.75 lumens per watt 



38 Louis DUNOYER 

(at a reduced operation of 1.45 amperes, the color temperature drops 
to 2250K. and the output to 6.70 lumens per watt). At an operating 
current of 5.5 amperes the color temperature of the spiral filament 
of the apparatus with slit is 2570K. on the internal surfaces of the 
turns and 2400K. on the external surfaces, the average output 
being 15.4 lumens per watt. The lamp of the apparatus with slit, 
therefore, is much more powerful than ours; its light is richer in 
blue rays to which the cell is more sensitive. If our filament were 
brought to the same temperature, although at the expense of its 
duration, the efficiency of the apparatus measured as above would 
be practically doubled since the output of the filament would be 
increased from 7.75 to 15.4 lumens per watt. 



THE RAPID RECORD OSCILLOGRAPH IN SOUND PICTURE 

STUDIES* 



A. M. CURTIS, T. E. SHEA, AND C. H. RUMPEL** 



Summary. This paper describes a special oscillograph which was designed for 
making rapid records in sound picture studies. The oscillograph is briefly described, 
and illustrations are presented of records obtained in making the following studies: 
microphonic action of vacuum tubes; noise levels in amplifiers; investigations on 
rectifiers; studies on light valve clash; action of the biasing current of light valves 
as used in noiseless recording by the variable density method; acoustical studies 
showing the rise and decay of transients; loud speaker selection with regard to load 
carrying capacity and mechanical flutter investigations of reproducer sets. 

The recording oscillograph, although an extremely valuable in- 
strument, is not in general very popular with engineers. This is 
especially true in sound picture work where time is often limited 
and the minutes and sometimes hours which must elapse after the 
oscillogram is taken and before it can be readily examined are a 
serious drawback. In addition, most types of recording oscillographs 
are found to be so insensitive over the frequency band used in sound 
pictures that the information which they give is frequently unreliable. 

About two years ago the Bell Telephone Laboratories, realizing the 
limitations of the available oscillographs of the recording type, under- 
took to design an instrument which would as far as possible avoid 
these shortcomings. The instrument which was evolved is capable 
of recording frequencies accurately up to 6000 cycles per second, 
and can furnish a developed record almost immediately after the 
oscillogram has been taken. Usually, therefore, oscillograms may be 
taken as rapidly as the conditions under investigation can be changed, 
and the results of the changes may be known at once. 

The oscillograph illustrated in Figs. 1 to 4 may be divided into 
two main parts, the galvanometer and the photographic mechanism. 
The galvanometer is of the string type, and is not unlike the light 
valve familiar to most sound picture engineers. There are, however, 

* Presented at the Spring, 1931, Meeting at Hollywood, Calif. 
** Bell Telephone Laboratories, New York, N. Y. 

39 



40 



CURTIS, SHEA, AND RUMPEL 



[J. S. M. P. E. 



three independent strings which permit the observation and record- 
ing of three simultaneous and separate phenomena with their phase 
relations. 




FIG. 1. Front view of the oscillograph. 



A tungsten filament lamp and a simple lens arrangement magnifies 
the motions of the strings and compresses their shadows to black spots 
on a line of light which extends across the 35-mm. bromide recording 
paper. This provides an oscillogram with white lines on a dark 



Jan., 1932] 



RAPID RECORD OSCILLOGRAPH 



41 



gray background. Means are also provided to photograph amplitude 
and timing lines on the oscillogram if desired. 

The photographic mechanism takes care of the exposing, develop- 
ing, and fixing of the oscillogram. The exposing is done by passing the 
paper through the line of light at the desired speed, using a system of 
rollers rotated by the exposure motor. The paper, having been 




FIG. 2. 



Front view of the oscillograph with covers removed to show part 
of the exposing and developing mechanism. 



exposed, is passed down a small chute into the developer tank through 
which it slowly travels by means of conveyor belts. From the 
developer the oscillogram is led into the fixing bath and is then 
passed out into a large fixing tank where it may be observed. Since 
oscillograms are generally taken much more rapidly than they are 
developed, a storage tank is provided into which the excess of ex- 



42 CURTIS, SHEA, AND RUMPEL [J. S. M. P. E. 

posed paper is passed, where it remains until led through the de- 
veloper. 

The process of taking an oscillogram is briefly as follows: the 
sources of current which it is desired to investigate are connected to 




FIG. 3. Rear view of the oscillograph. 

the galvanometer strings and the controls are adjusted until suitable 
deflections are obseryed on the viewing screen of the camera. The 
motors are then started and an operating lever is pulled out. After 
the deflection of the string images on the screen show that the ex- 



Jan., 1932] 



RAPID RECORD OSCILLOGRAPH 



43 



pected phenomenon has occurred, the operating lever is returned to 
the normal position. The developed and fixed oscillogram begins to 
pass before the operator's view about ten seconds later. It may then 
be examined immediately, measured, and, if desired as a permanent 




FIG. 4. Rear view of the oscillograph taken from another angle. 



record, returned to the large hypo tank for a few minutes to complete 
the processing. 

A particular feature of this instrument is the sharp definition of the 
string image, permitting accurate observations to be made with 



44 



CURTIS, SHEA, AND RUMPEL 



[J. S. M. P. E. 



deflections much smaller than are common with other types of 
recording oscillographs. This allows the oscillograms to be enlarged 
for analysis, and tracks having a height as great as four inches still 
give sharply defined lines. 

In order that the use of such an oscillograph in sound picture 
studies may be illustrated, a number of oscillograms have been 
prepared showing the application of this oscillograph to the solution 
of problems which are continually being investigated so that the 
sound picture may attain a greater degree of excellence. These 




0, 45 



0.50 



1,25 1,30 2,45 

TIME IN SECONDS FROM rMPACT- 



2.50 



FIG. 5. Oscillogram of microphonic response of vacuum tubes. 



oscillograms are not intended to show complete results of the various 
investigations but rather to point out how effectively this instrument 
may be used. 

1. Microphonic Vacuum Tube Studies. Microphonic vacuum 
tubes have imposed certain limitations both in recording and in 
reproducing systems. Those tubes, commonly used at low levels 
because of operating limitations, have been more microphonic than 
the higher powered tubes. Fig. 5 shows an oscillogram taken to 
illustrate the improvement which has been made in the microphonic 



Jan., 1932] 



RAPID RECORD OSCILLOGRAPH 



45 



response of a small vacuum tube as the result of studies which have 
been carried on during the past year. In this oscillogram three tubes 
were placed at the input to three amplifier channels having the same 
gain, each channel terminating at one of the oscillograph strings. 
The mounting upon which the three tubes were placed was given a 
single rap, causing the microphonic response of the tubes as shown. 
The relative freedom from microphonic effects of a recently produced 
vacuum tube (Tube B) is easily seen from a comparison with the 



A OVERLOADED WITH FILTER 




FIG. 6. Oscillograms of single tube amplifier. 



response of the earlier type of tubes recorded on the two outer strings 
(Tubes A and C). 

2. Amplifier Studies. In recording, it is common practice to 
operate a large number of recording amplifiers from a common "B" 
battery. Figs. 6 and 7 show the noise level across this battery due to 
a single amplifier, which may cause objectionable cross- talk in the 
other amplifiers unless precautions are taken to reduce the effect. 
Fig. 6 illustrates the effect in a single tube amplifier operated both 
within its rating and overloaded; Fig. 7 shows the corresponding 



46 



CURTIS, SHEA, AND RUMPEL [J. S. M. P. E. 




C WITHOUT FILTER A WITH FILTER 

OVERLOADED 



FIG. 7. Oscillograms of push-pull amplifier. 





B BATT. 
FIG. 8. Oscillogram of amplifier blocking. 



Jan., 1932] 



RAPID RECORD OSCILLOGRAPH 



47 




TOP BOTTOM 
STRING STRING 




FIG. 9. Oscillograms of rectifier characteristics. 




FIG. 10. Oscillograms of rectifier flickering. 



48 



CURTIS, SHEA, AND RUMPEL 



[J. S. M. P. E. 



effects in a push-pull amplifier. In each case record A illustrates the 
amplifier equipped with a simple filter composed of a single condenser 
and a small inductance, and indicates the absence of noise at the 
battery terminals. Records B and C show the presence of noise 
when the filter is removed. It is particularly interesting to note the 
presence of noise in the case of the push-pull amplifier as this is con- 
trary to the fairly common supposition that such an amplifier is 
totally without this effect. The tubes used in obtaining this record 
were carefully selected for equivalence. 







FIG. 11. Oscillograms showing light valve clash. 

In a-c. operated amplifiers blocking is a familiar phenomenon 
due frequently to a common plate impedance between the various 
tubes. In certain cases, however, it is very difficult to determine 
what is taking place. Fig. 8 shows a simple form of blocking in a 
two-stage amplifier. A small change in the plate current in tube 2 
causes a change in the plate current of tube 1 through the common 
plate impedance RI. This resistance R\ represents the impedance 
of the plate power supply which, in the case of an a-c. rectifier or a 
run-down "B" battery, might be quite high. 

3. Rectifier Investigations. Investigations of rectifier characteris- 



Jan., 1932) RAPID RECORD OSCILLOGRAPH 49 

tics may be readily made with this oscillograph as is shown in Figs. 
9 and 10. Record A , Fig. 9, shows the effect of working a gas rec- 
tifier tube directly into a capacity, as compared with working into an 
inductance (record B). This record was taken to determine the 
magnitude of the current peaks for the two types of filter, in order to 
determine the optimum condition for tube and condenser operation 
and thereby assure maximum service from the equipment. 

Fig. 10 shows a peculiar phenomenon found in certain full- wave 
gas rectifier tubes when operating under light loads. The tubes may 
be seen to flicker at various frequencies, as shown in the oscillogram. 
In each case the upper trace is the input voltage of the filter, the 
middle and lower traces the voltage between each plate and the 
filament of the rectifier tube. 

4. Light Valve Studies. Figs. 11, 12, and 13 show how the 
oscillograph may be used in studying the light valve used in the 
variable density method of recording. In Fig. 11 light valve string 
clash is shown. It may be seen how, while the bus voltage (input to 
the power amplifier feeding the valve) and valve voltage are un- 
affected, the output, as picked up by a monitoring photoelectric cell 
placed back of the film, is considerably distorted, but on one side of 
the cycle only. 

Figs. 12 and 13 illustrate the action of the biasing current of 
the light valve as used in noiseless recording by the variable density 
method. Fig. 12, record A, shows the action which takes place dur- 
ing attack or beginning of a sound wave. During this time the bias 
as shown by the center trace is being removed, allowing the light 
valve strings to resume their normal average spacing. This par- 
ticular condition illustrates the effect of too slow a removal of the 
biasing current causing the valve ribbons to clash, as indicated by the 
irregularities in the wave shown on the bottom trace. Record B is 
the same as record A except that speech is used to modulate the 
valve instead of a single frequency. Fig. 13 shows the decay after the 
input to the valve has died down. It may be seen that the bias is 
placed upon the valve much more slowly than it is removed in order 
that the low level portions at the ends of the various sounds as they 
die down will not be cut off. The reason for this may be seen from 
Figs. 14 and 15. 

5. Acoustical Studies. Figs. 14 and 15 are records of >sound 
build-up and decay, and were taken by placing the input to a loud 
speaker on the middle string or trace and picking up the sound by 



50 



CURTIS, SHEA, AND RUMPEL [j. s. M. p. E. 



A SINGLE FREQUENCY 400^ 




VALVE VOLTAGE 
,BUS VOLTAGE 



MONITOR VOLTAGE 



B SPEECH (SPOKEN NO* 4) 



FIG. 12. Oscillograms of noiseless recording light valve bias (attack). 



BUS VOLTAGE 



CURRENT 



MONITOR VOLTAGE 



A SINGLE FREQUENCY 400^ 



fUS VOLTAGE^ / VALVE VOLTAGE 



^MONITOR VOLTAGE 



B SPEECH (SPOKEN NO. 4: 



FIG. 13. Oscillograms of noiseless recording light valve bias (decay). 



Jan., 1932] 



RAPID RECORD OSCILLOGRAPH 



51 




FIG. 14. Oscillogram of sound growth. 



B CONTINUATION OF RECORC 




FIG. 15. Oscillogram of sound decay. 



52 CURTIS, SHEA, AND RUMPEL [J. s. M. P. E. 

means of two differently placed microphones whose amplified outputs 
are shown on the outer two traces. The sound growth curve of Fig. 
14 shows that in the case of a single frequency the sound builds up 
to its normal value very rapidly but may then drop or rise slightly 
depending upon the position of the microphone and the interference 
patterns set up. Fig. 15 illustrates different ways in which sound may 





FIG. 16. Oscillogram of loud speaker overload. 

decay when the input to the loud speaker is cut off. These differences 
are the results of interference, and it will be noted that they occur 
during the interval immediately following the cut-off of energy to 
the loud speaker. Being transient phenomena, the oscillograph 
is well suited to study them and is particularly valuable as an instru- 
ment supplemental to the reverberation meter, as by means of it 




FIG. 17. Oscillogram of mechanical flutter. 

particular portions of a decay curve may be studied in detail. How- 
ever, because of the limited amplitude of the oscillograph record, 
over-all reverberation times are most accurately measured by means 
of a reverberation meter. 

In selecting a loud speaker for a particular application it too fre- 
quently happens that this selection is made on the basis of frequency 



Jan., 1932] RAPID RECORD OSCILLOGRAPH 53 

response characteristics, and that the load carrying capacity is en- 
tirely neglected. The record of Fig. 16 was taken to illustrate this 
point. The upper and lower traces of this record indicate the sound 
outputs, as picked up by two similarly located microphones, of two 
loud speakers each receiving the same input. Speaker 1 is obviously 
overloaded. 

6. Mechanical Flutter Investigations. Fig. 17 shows how the 
oscillograph may be used to assist in mechanical design. Three sepa- 
rate sound film reproducers were set up, and the output of each 
reproducer was put on one string of the oscillograph. The same 
sound print of a thousand cycle film record was used on each of the 
reproducers with the results as shown. The upper trace is the output 
of a normal reproducer. The middle trace shows a reproducer having 
too large a driving sprocket at the sound gate. The lower trace 
shows the effect of having the driving sprocket slightly eccentric 
producing ninety-six and six cycle flutter. 



PHOTOGRAPHIC SENSITOMETRY, PART III* 
LOYD A. JONES** 



Due to its length, Mr. Jones' paper on sensitometry which was presented in part 
on three consecutive days at the Spring, 1931, Meeting of the Society at Hollywood, 
Calif., will be published in the JOURNAL in four issues. The following is the third 
of the four installments. The paper deals in a tutorial manner with the general 
subject of sensitometry, its theory and practice. The fourth installment will be 
published in the March, 1932, issue of the JOURNAL. 

OUTLINE 

I. Introduction. 

(A) Definition. 

(B) Scope of field. 

(C) Applications. 

CD) The characteristic D-log E curve. 

II. Sensitometers. 

(.4) Light sources. 

(1) Historical resume. 

(a) Natural light (sunlight, skylight, etc.). 

(b) Activated phosphorescent plate. 

(c) British standard candle. 

(d) The Hefner lamp. 

(e) The Harcourt pentane standard. 
(/) The acetylene flame. 

(g) Electric incandescent lamps. 

(2) Spectral composition of radiation. 

(a) The spectral emission curve. 
(6) The complete radiator. 

(c) Color temperature of sources. 

(d) Effect of color temperature on sensitivity values 

(3) Modern standards of intensity and quality. 

(a) Acetylene flame plus dyed gelatin filter. 
(6) Acetylene flame plus colored glass filter. 

(c) Acetylene flame plus colored liquid filter. 

(d) Electric incandescent, plus colored filters 

(4) The international unit of photographic intensity. 
(B) Exposure modulators. 

(1) Intensity scale instruments. 

* Presented at the Spring, 1931, Meeting at Hollywood, Calif. 
** Kodak Research Laboratories, Eastman Kodak Co., Rochester. N. Y. 
54 



PHOTOGRAPHIC SENSITOMETRY 55 

(a) Step tablets (/ variable by finite increments). 
(6) Wedge tablets (/ variable by infinitesimal incre- 
ments). 

(c) Luther's crossed wedge tablet. 

(d) Tube sensitometer. 

(e) Optical systems with step diaphragms. 

(/) Optical systems with continuously variable dia- 
phragms. 
(2) Time scale instruments. 

(a) Exposure intermittent. 

Finite exposure steps (discontinuous gradations). 
Infinitesimal exposure steps (continuous grada- 
tions). 
(6) Exposure non-intermittent. 

Finite exposure steps (discontinuous gradations). 
Infinitesimal exposure steps (continuous grada- 
tions). 

III. Development. 

04) Developers. 

(1) Standards for sensitometry. 

(a) Ferrous oxalate. 

(b) Pyro-soda. 

(c) ^-Aminophenol. 

(2) Standards for control of processing operations. 
OB) Temperature control. 

(C) Development technic. 

(1) For standardized sensitometry. 

(2) For control of processing operations. 

IV. The measurement of density. 

04) Optical characteristics of the image. 

(1) Partial scattering of transmitted light. 

(2) Diffuse density. 

(3) Specular density. 

(4) Intermediate density. 

(5) Relation between diffuse and specular values. 

(6) Effective density for contact printing. 

(7) Effective density for projection. 

(8) Color index. 

(B) Fog and fog correction. 

(1) Source of fog. 

(a) Inherent fog. 
(&) Processing fog. 

(2) Fog correction formulas. 

(C) Densitometers. 

(1) Bench photometer, 
(a) Rumford. 
(&) Bunsen. 



56 LOYD A. JONES [J. s. M. p. E. 

(c) Lumer Brodhun. 

(2) Martens polarization photometer. 

(a) Simple illuminator. 

(&) Split beam illuminator. 

(3) Integrating sphere. 

(a) For diffuse density. 

(b) For diffuse and specular density. 

(4) Completely diffused illumination. 

(a) For diffuse density. 

(5) Specialized forms. 

(a) Furgeson, Ren wick, and Benson. 

(b) Capstaff-Green. 

(c) High-intensity (Jones). 

(d) Density comparators. 

(6) Physical densitometers. 

(a) Thermoelectric. 

(b) Photoelectric. 

(c) Photovoltaic. 

V. Interpretation of Results. 

04) Speed or sensitivity. 

(1) Threshold speed. 

(a) Scheiner speed numbers. 

(b) Eder-Hecht. 

(2) Inertia speeds. 

(a) H & D scale. 

(b) Watkins scale. 

(c) Wynne scale. 

(3) Luther's crossed wedge method. 

(4) Minimum useful gradient. 

(B) Gamma infinity, 7^. 

(C) Velocity constant of development, K. 

(D) Time of development for specified gamma. 

(1) T d (y = 1.0). 

(E) Latitude, L. 

(F) Fog, F. 

VI. Spectral Sensitivity. 

(A) Dispersed radiation methods. 

(1) Monochromatic sensitometers. 

(2) Spectrographs. 

(a) Ordinary. 

(b) Glass wedge. 

(c) Optical wedge. 

(B) Selective absorption methods. 

(1) Tricolor. 

(2) Monochromatic filters. 

(3) Progressive cut filters. 



Jan., 1932] 



PHOTOGRAPHIC SENSITOMETRY 



57 



V. INTERPRETATION OF RESULTS 

Having now exposed the photographic material to a definitely 
known quantity and quality of radiation, developed the exposed 
material under standardized conditions, and measured the densities 
resulting from the various exposures, it remains to interpret the 
results thus obtained. As previously stated, some sensitometric 
testing methods, such, for instance, as the Scheiner, Eder-Hecht, 
etc., do not require the measurement of density, the result being 
judged directly by inspection of the developed material. Much 



28 



24 



20 



1.6 



I-Z 






^4 



OA 



0. I 1.1 f t.Q, 2.0 ZA Z.W 

x tA L.OG. EXPOSURE: (r-vcs} 



FIG. 38. Typical curve showing the relation between density and log 

exposure. 



more complete information may be obtained by methods involving 
the measurement of density and subsequent analysis of the results. 
For this purpose it is customary first to express the results in graphic 
form, and then to read off directly or to compute, by means of pre- 
viously established analytical relationships, the values of certain 
numerical constants useful in specifying the characteristics of a 
photographic material. The various graphic forms in which the 
sensitometric data may be shown will now be considered, after which 
numerical values derived therefrom and their significance for various 
theoretical and practical purposes will be considered. 



58 LOYD A. JONES [J. S. M. p. E. 

By plotting density, D, as a function of the logarithm (to the base 
10) of the exposure, logio E, a curve as shown in Fig. 38 is obtained. 
This is the graphic form proposed first by Hurter and Driffield (loc. 
cit.) for the presentation of sensitometric data and is therefore quite 
commonly referred to as the H & D curve although the terms D-log E 
curve and characteristic curve are frequently used in reference 
thereto. It has been found experimentally that in the case of many 
photographic materials a considerable portion of the D-log R curve 
is represented satisfactorily, within the limits of experimental errors, 
by a straight line. The limits of the straight line region are desig- 
nated by the points, A and B. The exposure region covered by the 
straight line portion of the characteristic curve is the region of correct 
exposure since throughout this exposure range density is directly 
proportional to log E. Therefore, for the correct proportional 
rendering of the negative of the various object brightnesses, the 
camera exposure must be adjusted so that only the straight line 
region is used. For the fulfillment of this condition the minimum 
density in the negative (corresponding to the deepest shadow in the 
object) must not be less than that of point A and the maximum 
negative density (corresponding to the highest light in the object) 
must not exceed that of point B. 

The relation between a given log E interval or increment, A log E, 
and the corresponding density interval or increment, AD, is given by 
the ratio AD/ A log E which is an expression of the average slope or 
gradient, G, for the interval A log E. Since the gradient is not in 
general constant, but changes continuously from point to point (for 
instance, in the region between C and A), it is necessary in order to 
express the gradient of the curve at any point to reduce the finite 
increments A log E and AD to the corresponding infinitesimal incre- 
ments d log E and dD. The gradient, G, at any point reduces 
therefore to the differential form 

G = dD/d log E 

For the straight line portion, however, G is constant and may be 
conveniently expressed in terms of the angle a subtended by the line 
AB and the log E axis. The tangent of this angle is called gamma, 7. 
For the straight line portion, therefore, 

G = dD/d log E = constant = tan a y 
Thus, gamma is the proportionality factor giving the relation between 



Jan., 1932] - PHOTOGRAPHIC SENSITOMETRY 59 

a given log E difference, A log E, and the corresponding density 
difference, AD. 

AD/ A log E = y 

Thus, if, in an object being photographed, two areas have brightnesses 
of 10 and 80 units, the A log value becomes log 80 - log 10 = 0.90. 
Now, if both are rendered on the straight line portion of the D-log E 
curve and if 7 = 0.8, then 

AD/0.90 = 0.8 
and 

AD = 0.72 

If a = 45 degrees, tan a or 7 becomes unity and any log E increment is 
rendered in the negative by an identical density difference. This is 
the condition which must be fulfilled if it is desired to reproduce 
exactly in the negative the brightness contrast in the object. If 
gamma is less than unity, correct proportional reproduction will be 
obtained but with compression of the brightness scale, while, if gamma 
is greater than unity, correct proportional reproduction will also be 
obtained but with expansion of the object brightness scale. 

Since gamma is equal to the ratio of the negative density difference 
to the corresponding log exposure difference, it is frequently used as a 
means of expressing the contrast of the negative or of the photographic 
material. It should be borne in mind constantly that gamma gives 
information pertaining only to the straight line portion of the curve 
and tells nothing of the contrast characteristics of other portions of 
the D-log E curve. This sensitometric constant is of great value and 
importance in both the theory and practice of photographic sensi- 
tometry. 

Projection of the straight line portion of the D-\og E curve on the 
log E axis determines the log exposure range over which direct 
proportionality between D and log E exists. By dropping perpendicu- 
lars from A and B to the log E axis the points M and TV are established. 
These fix the limits of this exposure range. The distance between 
M and N is called latitude, L, and may be expressed either in log 
E units or in exposure units. Thus, 

Latitude, L = log E n log E m (Log E units) 
or 

Latitude, L = E n /E m (Exposure units) 

The value of latitude for any given D-log E curve determines the 
maximum object contrast (ratio of maximum to minimum object 



60 LOYD A. JONES [J. s. M. p. E. 

brightness) which may be rendered with strict proportionality be- 
tween density and log exposure on that photographic material 
processed under the specified conditions used in obtaining the charac- 
teristic curve. Latitude is not a constant for a given photographic 
material, since its value depends profoundly upon the extent to which 
development is carried and, to a lesser extent, on other processing 
factors. It depends also upon certain exposure conditions, such as 
the quality (spectral composition) of the exposing radiation. 

The straight line, AB, extended cuts the log E axis at the point x 
and the value of E at this point is called the inertia, i. Since a 
material of low sensitivity has a high inertia value, and vice versa, it is 
necessary to take the reciprocal of the inertia in order to obtain a 
value which is directly proportional to sensitivity, hence 

Sensitivity oo l/i 

The absolute values obtained by taking the reciprocal of the inertia 
may be inconvenient for practical purposes since they may be less 
than unity, and hence expressible only as decimals or fractions. It is 
customary, therefore, in setting up practical sensitivity or speed 
scales to multiply this reciprocal by a constant, k, chosen more or less 
arbitrarily so as to give a series of convenient numbers. In general, 
therefore, speed is defined by the equation 

Speed, S = l.k 

The values of k commonly used will be discussed later. 

Now, from point A (Fig. 38) the ZMog E curve continues to the left 
into the region of decreasing exposure with constantly decreasing 
gradient, G, until at the point C this gradient becomes zero (G = 0), 
that is, the curve becomes parallel to or, if proper correction for fog 
has been made, coincident with the log E axis. This region, C to A , 
is called the region of underexposure or sometimes the toe of the 
characteristic curve. vSince the gradient, dD/d log E, decreases 
progressively from A to C, it follows that the density difference, AD, 
corresponding to a given small A log E, decreases continuously as the 
exposure is decreased, becoming zero at the exposure value correspond- 
ing to the point C. Thus, the power of the photographic material to 
show detail due to brightness differences in the object becomes less 
and less throughout the underexposure region vanishing entirely at an 
exposure value corresponding to the point C. 

From the point B, the upper limit of the straight line, the curve 



Jan., 1932] 



PHOTOGRAPHIC SENSITOMETRY 



61 



continues to the right into the region of increasing exposure with a 
constantly decreasing gradient until at the point D the gradient 
becomes zero, that is, the curve becomes parallel to the log E axis. 
The value of density corresponding to the point D is the maximum 
density, D max , obtainable with the specified processing conditions, 
development time, developer constitution, temperature, etc. Its 
value is not fixed entirely by these processing factors, but depends to 
some extent upon the quality of the exposing radiation. This region, 
B to D, is called the region of over exposure, or sometimes the shoulder 
of the characteristic curve. Here, as in the underexposure region, the 



Z4 



z.o 



1.2 



0.8 




0.4 



\ 



OA o.a 



1.7. I. ft 

L.OC.. 



Z.O 



X L.OC.. E-XPOSURe. (M.C.ft.^ 

FIG. 39. D-log E curves obtained with development times of T and 2T. 



density difference, AZ2, corresponding to a small log difference A log 
E, decreases progressively with increasing exposure and becomes zero 
at point D. Thus, the detail rendering power decreases progressively 
with increasing exposure and vanishes completely at the exposure 
value corresponding to point D. 

Points C and D, therefore, represent the limits of the exposure range 
within which the material is capable of rendering an object brightness 
difference by some density difference although near the limits 
(points C and D) this may be negligibly small, even for very great 
object brightness differences. This exposure range is termed the 



62 



LOYD A. JONES 



fj. S. M. P. E. 



total scale of the material and may be expressed either in log E units or 
as the ratio of the limiting exposures. Thus, 

Total scale = log EL log EC (Log E units) 
or 

Total scale = E L /E C (Exposure units) 

The latter form is more commonly used and is perhaps better for most 
purposes, since it is more directly interpretable in terms of the ratio of 
maximum object brightness to minimum object brightness, which is 
the form in which data relative to the brightness of the object are 
usually available. 

The shape, and frequently the position of the characteristic curve, 



!* 



o.e> 



04 





z e> 



%<b 



FIG. 40. Family of Z>-log E curves illustrating the approach to 7 oo for in- 
creasing development time. 

depends upon the development conditions. A simple case is illus- 
trated in Fig. 39 in which curve No. 1 represents the P-log E charac- 
teristic obtained for a development time of T, while curve No. 2 is 
that obtained for a development time of 2T. In this particular case 
the straight lines intersect at a point x lying on the log E axis. There- 
fore, the inertia value is the same for both times of development, and 
speed expressed in terms of inertia is the same for both curves. This is 
by no means true for all materials and all processing conditions, since 
in many cases the intersection point of the straight line portions is 



Jan., 1932] PHOTOGRAPHIC SENSITOMETRY 63 

found to lie either above or more frequently below the log E axis. 
Since inertia is defined as the value of exposure where the straight line 
extended cuts the log E axis, it follows that the value of the speed, 
based upon inertia, is not the same when determined from curves 
representing different developing times unless their intersection point 
lies on the log E axis. Angle a' is appreciably greater than a , hence 
gamma increases as development time is lengthened. This is true in 
practically all cases except when development is forced to such an 
extent that excessive fog is produced which may cause a decrease of 
gamma with development time. Such conditions are rarely met in 
practice and hence the statement that a increases with time of 
development is for all practical purposes a correct generalization. 
Projection of the straight line portion, A ' to B ', of curve No. 2 on the 
log E axis is appreciably shorter than that of the similar region, A to 
B, of curve No. 1, hence in this case latitude has decreased with the 
increasing time of development. The curves in Fig. 39 do not extend 
sufficiently far into the region of increasing exposure to show the final 
values of D max , but it is quite evident that the value of D max in- 
creases with development time. 

A somewhat more complete picture of the change in the shape of 
the D-\og E curve is shown in Fig. 40 in which the curves numbered 
from 1 to 6, inclusive, represent the data obtained from sensitometric 
strips developed for 2, 4, 6, 8, 10, and 12 minutes, respectively. 
Values of 7, AY, L, and i for these various times of development are 

TABLE XI 

Data Derived from Fig. 40 



Td 


y 


AT 


L 


i 


2 


0.50 




1.94 


0.10 


4 


0.80 


0.30 


1.74 


0.10 


6 


1.02 


0.22 


1.56 


0.10 


8 


1.14 


0.12 


1.40 


0.10 


10 


1.20 


0.06 


1.34 


0.10 


12 


1.24 


0.04 


1.30 


0.10 


a 


1.30 




... 


0.10 



shown in Table XI. For the two-minute development time a gamma 
of 0.50 is obtained. At 4 minutes gamma is equal to 0.8, an increase 
of 0.30. Increasing the time of development by another 2 minutes 
gives a gamma of 1.02, an increase of 0.22. For each successive two- 
minute addition to the development time the increase in gamma be- 



64 



LOYD A. JONES 



[J. S. M. P. E. 



comes less and less. This change in the rate of growth in gamma is 
more clearly shown in Fig. 41 (curve A) which is plotted from the 
data in Table XI. This curve is practically parallel to the T d axis at 
16 minutes and by extrapolation it is ascertained that it will not ex- 
ceed, for any kind of development, a value of 1.30. It is evident that 
as Td is prolonged, gamma approaches a limiting value, and this is 
called gamma infinity (TOO). This limiting gradient is illustrated in 
Fig. 40 by the dotted line designaed at 7. The value of gamma 
infinity is of great significance in both theoretical and practical sensi- 
tometry and will be discussed more fully a little later. 



10 



w 
<0.8 
I 



04 



/ / 
/ / 





FIG. 41. Time of development-gamma curves, A for high rate of develop- 
ment, B for low rate of development. Curve C is the corresponding time 
of development-fog curve. 

It will be noted by reference to Table XI and Fig. 40 that the value 
of latitude, L, tends to decrease as the time of development increases. 
Small vertical lines drawn through the various curves mark the limit 
of the straight line portions. In many cases when all of the straight 
lines (extended) intersect at a point lying on the log E axis, the points 
marking the limits of the straight line portions lie approximately on 
the circumference of circles drawn with the intersection point as a 
center. Under such conditions the actual length of the straight line 
is approximately constant, and hence is related in a definite manner to 



Jan., 1932] PHOTOGRAPHIC SENSITOMETRY 65 

gamma. The frequency of occurrence of this state of affairs is 
relatively low, and hence it is unsafe to attempt to make any 
generalization as to the relation between latitude and gamma except 
to say that latitude usually decreases as gamma increases, and, there- 
fore, as the time of development increases. 

Curves of the type shown in Fig. 41 are frequently of great value in 
analyzing the characteristics of a photographic material, particularly 
from the standpoint of its behavior during processing. These are 
known as time-gamma curves and are obtained by plotting gamma as a 
function of development time. As has been mentioned, curve A is 
obtained by plotting the data shown in Table XI which were derived 
from the family of characteristic curves shown in Fig. 40. Curve B 
illustrates the results obtained by processing the same material in a 
different developing solution. It is evident from a comparison of the 
two curves that gamma increases at a much lower rate in the case of 
curve B, although if the development time is sufficiently lengthened 
gamma appears to be approaching the same limiting value. 

The time-gamma curve is of use where it is desired to determine the 
development time which will yield some specified value of gamma. 
If such a curve is available for the material and the processing condi- 
tions being used, it is only necessary to read from the curve for any 
gamma value the corresponding development time. Such curves are 
also very useful in obtaining some idea as to the permissible variation 
in development time when it is desired to control processing so as to 
obtain gamma values lying within certain prescribed limits. For 
instance, let it be assumed that it is desirable to obtain a gamma of 0.6 
and that the permissible variations from the value are set at 0.03. 
The corresponding permissible variation in development time can be 
readily determined for the two conditions represented by curves 
A and B. The horizontal dotted lines are drawn through gamma 
values of 0.60 + 0.03 and 0.60 - 0.03. Where these horizontal lines 
intersect with curves A and B, perpendiculars are dropped onto 
the development time, T d , axis. In the case of curve A it is found 
that the development time must be held between 2.6 and 3.0 minutes, 
thus permitting a total allowable variation of 0.4 minutes which may 
be expressed as 2.8 0.2 minutes. In case of the curve B it is found 
for the same tolerance in gamma, minimum time is 5.7 and the maxi- 
mum 6.5, which may be expressed as 6.1 =*= 0.4 minutes. It is evident, 
therefore, that the allowable error in development time for the re- 
quired precision in control of gamma is twice as great in the case of 



66 



LOYD A. JONES 



[J. S. M. P. E. 



curve B as for curve A . The relation between a given gamma incre- 
ment and the corresponding development-time increment is, of course, 
given directly by the gradient of the y T curve at any particular 
point. If it is desired, therefore, to express numerically this relation- 
ship, it is only necessary to evaluate the differential dy/dt at any 
given point. The value of the differential at any point is inversely 
proportional to what may be termed processing latitude. In other 
words, the greater the gradient of the time-gamma curve at any point, 
the more precise must be the control of processing conditions in order 
to maintain a given tolerance in gamma. 




FIG. 42. Illustrating the general form of the first derivative, curve OA'B', 
of the D-log E curve, AB. 

Curve C in Fig. 41 shows the relation between fog and development 
time. Fog is determined by measuring the density of an area on the 
photographic material which has received no exposure but which has 
been developed. In general, for most photographic materials the 
value of fog is relatively low for the shorter times of development, but 
usually grows at an increasing rate as the development time is 
extended. Any value of fog which is given for a photographic 
material obviously must be accompanied by some specification of the 
development time or the extent of development (in terms of gamma) 
in order to have any definite significance. The complete Tj-fog 



Jan., 1932] 



PHOTOGRAPHIC SENSITOMETRY 



67 



curve is, of course, a complete representation of the relation between 
fog and the extent of development and in graphic methods of showing 
sensitometric results should be used rather than attempting to express 
this factor by a single numerical value. 

In Fig. 42 another useful graphic form is shown. Curve No. 1 
is the usual .D-log E characteristic curve. Curve No. 2 is the first 
derivative of the characteristic curve. It is obtained by plotting 
values of gradient, dD/d log E, as a function of log E. This curve 



3.0 



2.4 



.6 






30 



3.6 



1.8 ZA 

LOG E 

FIG. 43. Family of ZMog E curves obtained by plotting densities as read, 
without fog correction. 

shows somewhat more clearly the way in which gradient changes with 
log exposure. For the straight line portion of the curve lying between 
points A and B gradient is constant and equal to gamma. The first 
derivative curve throughout this region is a straight line parallel to 
the log E axis and having an ordinate value equivalent to gamma as 
shown on the gradient scale at the right of the figure. For values of 
exposure less than A and greater than B the first derivative curve takes 
the form as shown. This graphic form is useful where it is desired to 
determine precisely the exposure value corresponding to some 



68 



LOYD A. JONES 



[J. S. M. P. E. 



particular slope of the ZMog E curve. This form of presenting the 
data contains no more information nor can it be drawn with any 
greater precision than the D-log E curve itself, but for many purposes 
it presents the data in more convenient form and gives a more vivid 
mental picture of the relation between gradient and exposure. 

All the characteristic curves thus far shown have been plotted from 
data which have been corrected for fog. For many purposes for 
which sensitometric work is done this procedure is to be preferred, but 



1.4 



12 



1.0 











av 











<s> 







.4 



.3 



FIG. 44. The time of development-gamma curve derived from Fig. 43. 

for certain practical purposes it may be preferable to deal with the 
sensitometric data without making the correction for fog. This is 
true, for instance, in certain problems relating to tone reproduction 
where it is desired to obtain information as to actual density differences 
in the negative corresponding to known brightness differences in the 
object, and also to compute the time of exposure required for the 
making of a positive from the negative. In such problems it is 
essential to deal with the actual density values on the negative rather 
than with the corresponding values which have been corrected for fog. 



Jan., 1932] 



PHOTOGRAPHIC SENSITOMETRY 



69 



In Fig. 43 is shown a family of D-log E curves drawn from the measure- 
ments as read directly from the sensitometric strips without fog 
correction. It will be noted here that in the underexposure region 
the curves do not come down to the log E axis but become parallel to 
it at density values which are equivalent to the fog for the develop- 
ment times in question. In this group of curves it will also be noted 
that latitude decreases very markedly as the contrast or gamma of the 
characteristic curve increases. In Fig. 44 is shown the time-gamma 
curve plotted from values read from the curves in Fig. 43. This 




FIG. 45. The graphic representation of sensitometric characteristics of a 
high-speed negative material, including D-log E curves for various times 
of development, time of development-gamma curve, and time of develop- 
ment-fog curve. 



curve, of course, represents the effective contrast as a function of 
development time. It should be remembered that the correction for 
fog changes the values of the measured densities by different amounts, 
this change being proportionately greater for the lower densities, 
thus modifying the magnitude of gamma. When the data are to be 
used in tone reproduction problems, careful attention should be given 
to this point. 

Various ways of presenting sensitometric data in graphic form have 
now been considered and it is evident that in order to convey a maxi- 
mum of information more than one graphic form is necessary. It has 



70 



LOYD A. JONES 



[J. S. M. P. E. 



been found in practice that a complete family of .D-log E curves 
obtained with various development times together with a time-gamma 
curve and a time-fog curve serves as a fairly satisfactory graphic 




- z. o 



- \.o 



1.0 20 3.0 40 

FIG. 46. The graphic representation of sensitometric characteristics of 
motion picture positive film, including D-log E curves for various times 
of development, time of development-gamma curve, and time of development- 
fog curve. 

representation of the sensitometric characteristics. In Fig. 45 is 
illustrated one way in which these various functions may be con- 
veniently shown together. The characteristic curves themselves are 



Jan., 1932] PHOTOGRAPHIC SENSITOMETRY 71 

drawn in and the value of gamma for each is indicated. In the upper 
left-hand portion of the rectangle are shown the time-gamma and time- 
fog curves derived from the D-log E curves. The curves shown in 
Fig. 45 apply to a high-speed negative material for which gamma 
infinity is relatively low, being of the order of 1.4 to 1.5 as indicated 
by extrapolation of the time-gamma curve. In Fig. 46 is shown a 
similar group of curves for motion picture positive film which of course 
is a relatively slow, high contrast material. Having now dealt in 
some detail with the various graphic methods of presenting sensito- 
metric data, the problem of deriving from these graphic forms certain 
significant numerical values, which may be used as convenient 
specifications of these sensitometric characteristics, will be considered. 
It is quite impossible to express completely by means of a relatively 
few numerical values all of the information contained in the various 
possible graphic forms which may be used in presenting these data. 
Such numerical values, however, are convenient when it is desired to 
summarize in tabular form the sensitometric characteristics of various 
materials for purposes of record and intercomparison. 

SPEED OR SENSITIVITY 

In the case of negative materials one of the most important charac- 
teristics about which information is desired is that of sensitivity or 
speed, and in the earlier stages of the evolution of photographic 
sensitometry great emphasis was placed on the determination of this 
characteristic. Several different methods of expressing speed have 
been evolved and have been used rather widely in this country and 
abroad. It may be of interest to consider the significance of these 
various methods of speed specification and the inter-relation between 
the resultant numerical values. 

Threshold Speed. One of the earliest methods used for the express- 
ing of sensitivity was to specify the exposure required to produce a 
just perceptible density. In methods of sensitometry not involving 
the measurement of the developed densities this is the only feasible 
method of speed expression which can be used. This was adopted by 
Scheiner who devised a sensitometer which has already been described 
in an earlier section of this paper. The sector wheel in the Scheiner 
sensitometer was so cut that exposure increased logarithmically from 
1 to 100 units. The distance between the points on the photographic 
material corresponding to these exposure limits was divided into 
twenty equal steps, numbered consecutively from 1 to 20. The 



72 LOYD A. JONES [J. S. M. p. E. 

Scheiner speed scale, therefore, consists of numbers in arithmetic 
progression, 1, 2, 3, 4, etc., from 1 to 20, covering the sensitivity range 
of from 1 to 100. Relative sensitivity represented by any given 
number in the scale is 1.27 times as great as the relative sensitivity 
corresponding to the next lower number in the scale. This relation 
is shown in Table XII, in the first column of which the Scheiner 
numbers are given, and in the last column will be found the correspond- 
ing relative sensitivity values. The consecutive numbers of the 

TABLE XII 

F/209 Inter comparison of Speed Values as Expressed by Various Weil-Known 

Methods 



Scheiner 


Eder-Hecht 


H & D 


Watkins 


Wynne 


Relative 


1 


42 


7 


11 


F/21 


1.0 


2 


46 


9 


13 


F/24 


1.27 


3 


48 


12 


17 


F/21 


1.62 


4 


50 


15 


22 


F/30 


2.07 


5 


53 


19 


28 


F/34 


2.64 


6 


56 


24 


36 


F/38 


3.36 


7 


58 


31 


45 


F/43 


4.28 


8 


61 


40 


58 


F/49 


5.45 


9 


64 


50 


74 


F/55 


6.95 


10 


66 


64 


94 


F/63 


8.86 


11 


68 


82 


122 


F/71 


11.3 


12 


71 


104 


153 


F/79 


14.4 


13 


74 


133 


196 


F/90 


18.3 


14 


77 


170 


250 


F/101 


23.4 


15 


80 


216 


317 


F/114 


29.8 


16 


83 


276 


405 


F/129 


37.9 


17 


84 


351 


515 


F/145 


48.3 


18 


86 


448 


660 


F/165 


61.6 


19 


88 


570 


840 


F/196 


78.5 



20 90 727 1065 F/209 100.0 

Scheiner scale which increase in arithmetical progression correspond 
to a geometrical progression in relative sensitivity. The scale inter- 
val, therefore, is slightly greater than that given by using consecutive 
powers of the cube root of 2 in which the multiplying factor from step 
to step is the cube root of 2 or 1.26. 

The Eder-Hecht sensitometer, as has already been mentioned, is of 
the tablet type consisting of a neutral gray wedge with a continuous 
gradient. On this are printed a series of numbers in arithmetical 
progression and equally spaced. This speed scale is, therefore, also 



Jan., 1932] PHOTOGRAPHIC SENSITOMETRY 73 

of the logarithmic form, assuming that the neutral gray wedge has a 
constant gradient. The numbers actually used on the Eder-Hecht 
sensitometer tablet as compared with the Scheiner scale are as shown 
in the second column of Table XII. 

While the threshold method of expressing sensitivity has certain 
features to recommend it, it leaves much to be desired from the stand- 
point of precision and significance. The magnitude of the least 
perceptible density depends profoundly upon the conditions under 
which the inspection is made. In fact, the judgment which is actually 
made is not that of least perceptible density, but least perceptible 
density difference. Under the most favorable conditions of observa- 
tion, the human eye can detect a brightness difference of 1.7 per cent. 
This corresponds to a density difference of 0.008. Under other con- 
ditions of inspection, however, such as relatively low illumination and 
uncomfortable visual conditions, this just perceptible density differ- 
ence may be easily as great at 0.04. It is evident, therefore, that, 
unless great care is taken to standardize and maintain the visual condi- 
tions under which judgment of the just perceptible density is made, 
values of threshold speed read from the same actual test strip may fluc- 
tuate over a considerable range. Furthermore, if conditions are ad- 
justed to give maximum visual sensitivities so that a very slight density 
difference may be detected, such, for instance, as the value men- 
tioned above, namely, 0.008, the absolute value of speed is rather high, 
regarded from the practical standpoint. For instance, the point on 
the toe or underexposure region of the characteristic curve where a 
density of 0.007 is obtained, is in almost all cases at a point of ex- 
tremely low gradient. It is questionable whether the underexposure 
region at or near the point where D is equal to 0.008 is of any practical 
value. While it may be argued that the effective speed should be con- 
sidered to go down into the region of low exposures to the point where 
a just perceptible density is produced, this seems somewhat fallacious, 
when it is considered that the real function of a photographic material 
is to reproduce, as perceptible density differences, the brightness differ- 
ences which exist in the object. It seems, therefore, that we should 
be more concerned with the definition of speed in terms of the power 
of the material to reproduce satisfactorily some minimal contrast. 

Inertia Speeds. Hurter and Drirfield in their work on photographic 
sensitometry suggested that the speed of a material could be specified 
satisfactorily in terms of the inertia. They proposed, therefore, the 
expression of speed as the reciprocal of the inertia multiplied by a 



74 LOYD A. JONES [J. S. M. P. E. 

constant for which they chose a value of 34. The use of this number 
gave a series of speed values of convenient magnitude for practical 
use. In the third column of Table XII are shown the H & D speed 
numbers in direct comparison with relative values as shown in the last 
column. 

The Watkins speed scale is also based upon inertia, but instead of 
using 34 as suggested by Hurter and Driffield, Watkins adopted 68 
as the value of constant k. The actual relation between the Watkins 
and H & D numbers, however, indicates that the Watkins constant is 
more nearly 50 than 68. In the fourth column of Table XII are 
shown the values of the Watkins speed scale in comparison with the 
other well-known systems. 

The Wynne system of expressing speed is not used to any great 
extent but is of some interest. This is also based fundamentally upon 
inertia values but uses numbers which are expressed in terms of lens 
aperture as indicated by the symbol F which precedes the number. 
These numbers are proportional to the product of 6.4 by the square 
root of the Watkins number. A Watkins speed of 100 (equivalent to 
H & D speed of 68) gives a Wynne number F/64. This scale is shown 
in the fifth column of Table XII. 

For many purposes and under many conditions, the expression of 
speed in terms of inertia is of great value. As long as all of the 
straight line portions of a family of D-log E curves pass through a 
common intersection point and this point lies on the log E axis, 
inertia and hence speed are independent of development time. Under 
such conditions the speed becomes a very significant constant for 
the photographic material. Unfortunately the existence of a common 
intersection point lying upon the log E axis is frequently not found in 
practice. In most cases of normal development a common inter- 
section point is found, provided that proper corrections have been 
made for fog. This intersection point, however, very 'frequently 
lies below the log E axis and in relatively rare cases is located above 
that axis. This subject has been dealt with at great length by 
Nietz. 100 It has been found that in the presence of free bromide, 
whether it be in the developing solution or present in the photo- 
graphic material itself, the intersection point is, in general, depressed 
to a position below the log E axis. Such a condition is shown in Fig. 
47 which represents the straight line portions of a family of D-log E 
curves. Assuming for the moment that a common intersection 
point does exist, its coordinates may be represented by a and b as 



Jan., 1932] 



PHOTOGRAPHIC SENSITOMETRY 



75 



shown in Fig. 47, and it has been proposed to define the speed of the 
material in terms of the coordinates of this point. Under such 
conditions it is evident that the inertia is a function of gamma, and 
hence speed based upon inertia value will become a function of 
gamma, and a speed value of this nature can only be significant 
provided the corresponding gamma value is specified. For the 
purpose of certain theoretical investigations into the nature of 
exposure and development, a knowledge of the coordinates of the 




FIG. 47. The straight line portions of a family of ZMog E curves illus- 
trating the existence of a common intersection point below the log exposure 
axis, thus causing inertia to depend upon time of development. 

intersection point, as shown in Fig. 47, may be of great value, but it 
does not appear to be very significant for the purposes of determining 
the practical speeds. 

The extent of the dependence of the inertia speed upon gamma is 
illustrated in Table XIII. These data are derived from measure- 
ments made on a high-speed negative material processed in a develop- 
ing solution containing some bromide. The straight line portions of 
the P-log E curves intersect at a point well below the log E axis. The 



76 LOYD A. JONES [J. S. M. p. E. 

development times, T d , extend from 3 to 20 minutes, the correspond- 
ing gamma range being from 0.27 to 1.12. For this range in develop- 
ment, the reciprocal inertia changes from 22 to 140, a six-fold increase 
in speed as derived from inertia values. 

The complication involved in expressing speed by means of the 
inertia does not end here. Many materials are found for which there 
is no single point of intersection for the straight line portions of the 
D-log E characteristics. In fact, some materials do not give very 
satisfactory straight line relations between density and log 
exposure. There is wide divergence in the relative shape of the 
underexposure regions and, in fact, an almost endless variety of 
conditions are found which makes it extremely difficult to generalize 
satisfactorily the expression of practical or effective speed of photo- 
graphic materials for all purposes. Anomalous behavior in both 
shape and position of the D-log E curves resulting from development 

TABLE XIII 

Data Illustrating the Dependence of Inertia upon Time of Development 

Td y i l/i 



3 


0.27 


0.045 


22 


5 


0.43 


0.020 


50 


8 


0.68 


0.012 


82 


12 


0.83 


0.010 


100 


20 


1.12 


0.007 


140 



for different times seems to be particularly common in materials of 
high sensitivity. This subject has been discussed at considerable 
length by Sheppard. 101 He classifies emulsions generally into 
orthophotic and anorthophotic categories. Orthophotic materials 
show a definite convergence point of the straight line portion of the 
characteristic curves, while anorthophotic materials depart widely 
from this condition showing no tendency to give a common point of 
convergence. Sheppard concludes from his study of the subject that 
emulsions of the anorthophotic type have characteristics which are 
much less reproducible from batch to batch than those of the ortho- 
photic class. In many fields of work reproducibility is highly im- 
portant; for instance, when these materials are used as a means of 
making quantitative measurements in science and technology, when 
it is desired to employ automatic processing methods, and in those 
cases where it is desirable to make application of the laws of tone re- 
production. It would appear that this demand on the part of the 



Jan., 1932] PHOTOGRAPHIC SENSITOMETRY 77 

users of photographic material for reproducibility may tend auto- 
matically toward the rejection of materials of the anorthophotic type. 
Hence photographic materials may respond to evolutionary laws, 
their characteristics tending to become predominantly orthophotic as 
a result of the survival of the fittest. While it is impossible to ignore 
the existence of certain materials which, during the course of develop- 
ment, do not even approach to the classical behavior required by H & 
D theory, it would seem that in the development of sensitometry 
greatest attention should be paid to the evolution of sensitometric 
systems applicable particularly to the materials which do approach 
to normal types. The case, therefore, is not quite as hopeless as it 
may appear and it does not seem unreasonable to assume the existence 
of a normal type behavior from which the great majority of materials 
used in large volume depart but little and in a known and specifiable 
manner. In any case it seems more profitable to take the position 
that normality and orderliness are the rule rather than to assume the 
attitude of destructive criticism and maintain that "all photographic 
materials are exceptions," thus abandoning the entire field of syste- 
matic sensitometry to chaos. 

Luther's Crossed Wedge Method. In the section dealing with sensi- 
tometers the crossed wedge method proposed by Luther 23 for 
obtaining directly the D-log E curve is mentioned. A few words rela- 
tive to the interpretation of the results obtained in this manner seem to 
be in order. The envelope of the darkened area gives directly the 
D-log E characteristic. If the density and density gradient of each 
wedge are known, it is possible to establish the correct density and log 
exposure scales. The linearity of these scales, that is, the representa- 
tion of a specified density or log E difference by a constant linear 
interval throughout the respective scales, requires that the density 
gradient of the wedges be constant, and also that the wedge density 
be uniform along any line perpendicular to the gradient direction. 
Considering the difficulties involved in the manufacture of wedges of 
satisfactory uniformity, both of density and density gradient, having 
sufficient freedom from selective absorption, and in the exact 
calibration of these wedges, it does not seem likely that the precision 
obtainable in the final result can be as great as that resulting from 
the exposure of the photographic material in a well-designed and 
^carefully operated sensitometer of the time-scale type followed by the 
measurement of density with suitable densitometers. However, as a 
rapid and convenient means of testing, the method has much to 



78 LOYD A. JONES [J. S. M. P. E. 

commend it. Yielding as it does the typical D-log E curve, the 
interpretative methods are identical to those applicable to similar 
curves obtained by other methods, and values of the usual factors, 
such as gamma, inertia, fog, etc., may be derived. 

Minimum Useful Gradient. Thus far two methods for the expres- 
sion of sensitivity or speed have been discussed, one based upon 
exposure corresponding to a just perceptible density (Schwellenwert) , 
and the other based upon the value of inertia read at the point where 
the straight line portion of the D-log E curve cuts the log E axis. 
While both of these methods have certain points to commend them, 
both also have very serious deficiencies. Threshold speeds are not 
independent of the time of development and therefore cannot be said 
to be strictly a constant of the photographic material. Moreover, 
the absolute value obtained under inspection conditions yielding 
maximum visual sensitivity gives a speed which is too high, when it is 
desired to compute the exposure required for the satisfactory render- 
ing of detail in the shadow regions of the object. It has been seen, 
further, that speed values based upon inertia are dependent in many 
cases upon time of development and hence require an accompanying 
expression of gamma in order to be significant. Even under these 
conditions, the absolute value of speed is not very useful in computing 
the exposure time required in order to give satisfactory rendering of 
shadow detail. A study of a large number of characteristic curves 
shows that the gradient corresponding to the inertia value varies 
between wide limits. When it is considered that the chief function of 
a photographic negative material as used in practice is to reproduce as 
density differences the brightness differences existing in the object 
photographed, it seems logical to demand that the minimum useful 
exposure be determined by some specified gradient of the D-log E 
characteristic. This idea has been discussed by Luther 102 and its use 
advocated by him, especially in cases where the fog of the emulsion is 
relatively high. The subject has also been discussed at some length 
by Jones and Russell. 103 

There seems to be little doubt that this idea is based on a sound 
theoretical foundation. The difficulty met, however, is that of 
deciding upon the value which is to be taken as representing the 
minimum useful gradient. Luther suggested that this value should be 
0.5. Judging from data available in publications by Goldberg and 
found in a paper by Jones 104 dealing with the contrast of photographic 
printing papers, it appears that this value is too high and that satis- 



Jan., 1932] 



PHOTOGRAPHIC SENSITOMETRY 



79 



factory reproduction of object detail can be obtained by utilizing 
portions of the characteristic curve of lower gradient. The subject 
has also been discussed by Sheppard 101 who states that the minimum 
useful gradient "will in general depend not only upon the negative 
but also upon the positive aspect of tone reproduction so that its 
fixation is not expressible by a unique function of the negative 
material itself." This conclusion is undoubtedly correct and its 
validity is supported by the data and the discussion given by Jones 
(loc. tit.). 
It seems quite possible, however, for certain definite classes of work 



RADIENT 

& 5 




g 

<S> Ad 

I 
M 



LOG EXPOSURE 

FIG. 48. D-log E curve, O4J3, and its first derivative, FED. The con- 
struction illustrates the method of finding the relation between gradient 
and log exposure. 

to establish a value of minimum negative gradient in terms of which 
sensitivity or speed may be expressed in a manner of considerable 
practical utility. For instance, from a knowledge of common 
practice which results in acceptable tone reproduction in the field of 
motion picture photography, it is possible to draw fairly definite 
conclusions as to the minimum useful gradient of negative materials 
used in this work. This knowledge is based upon careful densi to- 
metric analyses of a large number of motion picture negatives and 
positives. The value indicated by the available information lies 



80 



LOYD A. JONES 



[J. S. M. P. E. 



between 0.2 and 0.3. A similar value is known to represent fairly well 
the conditions existing in the field of amateur photography. It is not 
possible with the information available at the present time to say 
definitely whether or not a fixed value of minimum limiting gradient 
can be chosen which would be satisfactory in all fields and, if so, just 
what the absolute value of this quantity should be. It is definitely 
known, however, that in the motion picture field and in the amateur 
field practically all negatives utilize the greater portion of the under- 
exposure region of the D-log E characteristic. Furthermore, it is 
logical to conclude that the toe of the characteristic curve below 




'3.4 3.0 l.<* 12. T.8 0.4 

LOG EXPOSURE 

FIG. 49. Typical P-log E characteristic curves illustrating the disagree- 
ment between speed values based upon inertia and those based upon minimal 
useful gradient. 

some definite value of gradient is too flat for satisfactory reproduction 
of object brightness differences. It seems desirable, therefore, to give 
this suggested method for the specification of speed very careful 
consideration and some data based upon an arbitrary assumption of 
the value of minimum limiting gradient may be of interest. For 
this purpose a value of G equal to 0.2 will be assumed. 

In Fig. 48 the most precise method of determining the exposure 
value corresponding to this specified gradient is illustrated. Curve 
OAB represents the underexposure and part of the correct exposure 
region of the .D-log E characteristic, curve FED being its first deriva- 
tive. Through the gradient value of 0.2 a horizontal line is drawn 



Jan., 1932] 



PHOTOGRAPHIC SENSITOMETRY 



81 



which establishes point D. A perpendicular dropped from this point 
cuts the characteristic curve at point O, which is the point having the 
gradient of 0.2. The exposure value, E m , corresponding to the point 
0, is that desired in order to express speed in terms of a minimum 
useful gradient equal to 02. Speed and sensitivity are of course 
inversely proportional to E m . 

A typical case which illustrates the merits of this method of express- 
ing speed is shown in Fig. 49. The two materials illustrated have 
been developed to the same gamma, and on the basis of the inertia 
method of expressing speed; the material of which the straight line 




35 



Z.3 



27 T.% 

LOG EXPOSURE 

FIG. 50. Family of D-\og E curves illustrating the dependence of minimum 
useful gradient speed upon time of development. 

portion cuts the log E axis farthest to the left has the higher speed of 
the two. The small arrows indicate the points on the underexposure 
region where G is equal to 02. It will be seen that the minimum 
limiting gradient method reverses the speeds of the two materials. 
If we express inertia speed as 10 times the reciprocal of the inertia 
and the gradient speed as 10 times the reciprocal of E m , the numerical 
results derived from Fig. 49 are as follows: 



220 



1360 
720 



82 LOYD A. JONES- [J. S. M. P. E. 

Further illustration of this suggested method is given in Fig. 50 
in which are shown the underexposure regions of four D-log E charac- 
teristic curves obtained by using different times of development. 
Again, the points of gradient equal to 0.2 are indicated by the small 
arrows attached to each curve. In the case of the shortest time of 
development it will be noted that the gradient for 0.2 is obtained for 

TABLE XIV 

Data Illustrating the Relation between the Values of Speed Based upon Inertia, 
i, and Those Based upon the Exposure, E m , Corresponding to the Minimum Useful 

Gradient 



Td 


7 


Fog 


i 


Em 


R 


1.5 


0.68 


0.15 


0.045 


0.0280 


1.6 


2.0 


0.80 


0.22 


0.025 


0.0058 


4.3 


4.5 


1.7 


0.32 


0.026 


0.0053 


4.9 


6.0 


2.0 


0.47 


0.024 


0.0044 


5.5 


9.0 


2.6 


0.51 


0.033 


0.0050 


6.6 



an exposure value practically equal to the inertia value, while for 
longer times of development the speed values based upon minimum 
gradient are very much higher than those based upon inertia. The 
data derived from the curves in Fig. 50 are shown in Table XIV. 
The ratio of i to E m is given in the last column. 

GAMMA INFINITY, y , AND CONSTANT OF DEVELOPMENT, K 

The importance of gamma infinity, both for theoretical and practical 
sensitometry, already has been emphasized and in Fig. 41 an experi- 
mental method of determining the value of gamma infinity has been 
illustrated. In certain cases, similar to that illustrated in Fig. 40 
where development takes place in the manner assumed by the classical 
H & D theory, it is quite possible to compute from the data contained 
in two or more J9-log E curves a theoretical value for gamma infinity. 
Along with this may be derived the value of K , the velocity constant of 
development. According to the method introduced by Sheppard and 
Mees, l05 this computation is based upon the gamma values obtained 
from two D-log E curves, for one of which the development time is 
twice that of the other. The necessary data are therefore derived 
from a pair of curves such as those shown in Fig. 39 where this ratio of 
development times was used. From theoretical considerations it can 
be shown that the relation between development time and growth of 
gamma can be expressed by the equation 

7 - 7.(1 - -*') (1) 



Jan., 1932] PHOTOGRAPHIC SENSITOMETRY 83 

By substituting in this equation values of gamma and time of develop- 
ment obtained from curve No. 1, equation (2) is obtained, and 
similarly by using values of gamma and the time of development read 
from curve No. 2, equation (3) is obtained. 

71 = 7^(1 - e-*0 (2) . 

72 = 7 00(1 - e-*} (3) 
Combining (2) and (3) we obtain 

Ti(l - -) = T2 (l - -*") (4) 

Since fe = 2fc 

I? = i + -, (5) 






(6) 



7i _ . 72 , R x 

n -Kt^ 1 ,,-R-fo V/ 



From the known values of 71, 72, /i, and / 2 it is therefore possible to 
compute values of 7 , and X. From any other pair of characteristic 
curves for which the times of development are related by the expres- 
sion /2 = 2/i, additional values of these constants may be computed 
which should, of course, check those based on any other similar pair 
of sensitometric curves. It should be emphasized that these 
theoretical relationships do not hold in all cases, their validity de- 
pending upon the normality of the family of D-log E characteristics 
as judged by the requirements of the H & D theory. 
By differentiation of equation (1) the relationship 
dy/dt = K(y m - 7) 

is obtained. Now dy/dt is the slope of a time-gamma curve such as 
that shown in Fig. 41. The values of this gradient can be determined 
graphically frpm an experimental time-gamma curve for any value of 
gamma, and if gamma infinity is known, having been determined, let 
us say, experimentally as illustrated in Fig. 41, it is possible to compute 
the corresponding value of K. All values of K computed in this way 
for various values of gamma should, of course, be the same provided 
the curve is of the exponential form. It is found in practice that 
when such a time-gamma curve is derived from a normal family of 
characteristic curves, and when the experimental determination of 
7 OP is valid, this condition, K equal to a constant, is fulfilled. In many 



84 LOYD A. JONES [J. S. M. P. E. 

cases, however, the value of K, computed from different assumed 
values of gamma and corresponding graphically determined values of 
dy/dt, is not constant. This is evidence of abnormality in the time 
of development-gamma relation and can usually be traced back 
and found to be due to improper correction for fog, lack of a common 
convergence point, or other departures from what may be termed the 
normal behavior. 

Certain development characteristics of any particular photographic 
material may be deduced from the values of 7 and K. 

For instance, if K is high and 7^ is high: 

Development will start quickly, proceed at a high rate, and gamma 
will continue to build up to a high value. Process plates and motion 
picture positive film are typical examples of the materials having 
these characteristics. 

If K is high and y m is low: 

The image will flash up quickly and 7 will build up rapidly at first 
but soon cease to increase, reaching a limit at a relatively low value. 
In the case of these materials the image appears very quickly but 
fails to carry through and build up high densities. 

If K is low and 7 ro is high: 

Development starts slowly and 7 increases at a relatively low 
rate, but by extending the time of development the value of 7 may be 
built up to a high value. 

If K is low and 7 is low: 

Development starts slowly, 7 increasing at a relatively low rate 
which very soon decreases and flattens out at a final low value. 
Further development will not serve to increase contrast. 

TIME OF DEVELOPMENT FOR A SPECIFIED GAMMA, T y = x 

In practical specification of sensitometric characteristics it is 
sometimes desirable to measure directly, in terms of a single constant, 
the dependence of gamma or contrast on time of development. Such 
a figure combines to a certain extent the information contained in the 
values of 700 and K and usually is more easily obtained. This is 
accomplished by stating the time of development required to give a 
specified gamma. In choosing the gamma value for which this time 
of development is stated, it is, of course, desirable to use the contrast 
to which the material is in practice usually developed. It is, of 
course, impossible to find any single value of gamma which fits the 
requirements of all possible classes of photographic work. In some 



Jan., 1932] PHOTOGRAPHIC SENSITOMETRY 85 

cases as, for instance, motion picture photography, the negative is 
usually developed to a relatively low gamma such as 0.5 or 0.6. In 
the amateur field somewhat higher gammas are usual. In portrait 
work it seems probable that a gamma of 0.8 represents a fair practice. 
In commercial work this value is unity or even somewhat above, 
while in process work gamma is pushed as near as possible to 7^, 
practical values lying between 1.8 and 3.0. In the case of motion 
picture positive film it is probable that a gamma of 1.8 represents a 
fair average. For the purposes of preparing tables showing relative 
sensitometric characteristics, a gamma value of unity is usually 
chosen, the time of development required by various materials to 
obtain this value under standardized processing conditions being 
determined. This factor is usually expressed symbolically as 
T y i.o. 

LATITUDE, L 

Latitude has already been defined in the earlier discussion and the 
method of determination explained. As stated previously, latitude 
varies with gamma and therefore only has significance when ac- 
companied by a statement in terms of gamma of the extent to which 
development has been carried. Various attempts have been made to 
find a theoretical or analytical expression relating latitude and gamma. 
While some equations have been proposed, none of these seem to be of 
sufficient general validity to warrant consideration. It is customary, 
therefore, to determine latitude graphically directly from the plotted 
characteristic curves. It is usual to express latitude in the form of 
the ratio of the maximum to the minimum exposure lying on the 
straight line portion of the curve. A knowledge of latitude is in 
certain classes of work of considerable importance. It defines the 
ratio of object brightnesses which may be rendered by the material, 
when developed to the gamma specified, without non-linear distortion 
of the object contrasts. For most high-speed negative materials, the 
magnitude of gamma is considerably greater than the ratio of maxi- 
mum to minimum brightness in average photographic subjects 
under normal illuminations. It is extremely unusual in out-of-door 
work to encounter subjects in which the contrasts, that is, ratio of 
maximum to minimum brightness, is greater than 100. Extreme cases 
show values as high as 250, but these are rare. In studio work it is, 
of course, possible to obtain artificial lighting in which the contrast is 
greater than that mentioned above. However, the measurement in a 
great many portrait and motion picture studios indicates that in this 



86 LOYD A. JONES [J. S. M. P. E. 

class of work contrast seldom exceeds 100 or at the most 200. For the 
low values of gamma usually used in motion picture studio and 
portrait work, it is not infrequent to find that the photographic 
materials have latitudes well above 500 and in some cases^reater 
than 1000. For the low-speed, high-contrast materials; gmSmfirbf 
course is much lower. Here again the value will depend upon the 
extent to which the material is developed. For the high contrast to 
which motion picture positive film is usually developed, a latitude of 
32 to 64 is usual. 

FOG, F 

The definition of fog and the method of measurement has already 
been defined in the previous discussion. Its value is dependent upon 
the extent to which development has been carried and of course is 
profoundly influenced by the composition of the developing solution. 
In giving fog as a sensitometric value it is necesaary, therefore, to 
specify both the composition of the developer used and the extent to 
which development has been carried usually in terms of gamma. 
As stated previously complete information as to fog giving propensi- 
ties of photographic material is shown best in graphic form by the 
time of development-fog curve as illustrated in Fig. 41. In cases 
where the time of development required to give gamma of unity and 
inertia for gamma of unity are given, it is customary to express fog 
also for gamma of unity. 

TABLE XV 

Sensitometric Constants of Typical Photographic Materials 

Td- 
Material F K y m (7 = 1.0) L * 10/t 

Motion picture film super- 
speed 0.15 0.24 1.6 4.0 400 0.010 1000 
Motion picture film normal 0.10 0.20 1.8 4.0 300 0.017 600 
Motion picture film positive 0.03 0.30 2.8 1.5 50 0.330 30 
Portrait film normal 0.08 0.20 1.6 5.0 200 0.020 500 
Portrait pan film super-speed 0.15 0.24 1.6 4.0 300 0.010 1000 
Amateur film, fast 0.10 0.16 1.8 5.0 150 0.017 600 
Amateur film, normal 0.07 0.18 1.6 5.5 80 0.025 400 
"Press" plate 0.15 0.16 1.8 5.0 100 0.010 1000 
Commercial ordinary 0.05 0.18 2.5 3.0 75 0.040 250 
Commercial ortho 0.08 0.19 2.2 3.2 75 0.028 350 
Commercial pan 0.10 ' 0.20 2.0 3.5 100 0.020 500 
Process plate ordinary . 04 . 28 3.0 1.5 25 . 250 40 
Process plate pan 0.08 0.28 3.0 1.5 25 0.067 150 
Lantern plate 0.03 0.32 3.2 1.2 25 0.650 15 



S 



Jan., 1932] PHOTOGRAPHIC $ENSITOMETRY 87 

In Table XV are shown some typical numerical sensitometric 
values for a variety of photographic materials. The values given do 
not refer to any particular material but represent a fair average of 
materials which fall within the classifications as indicated in the 
first column. They neither represent the best now available nor the 
"ideal" material in each class, but specify the characteristics that may 
reasonably be expected of these materials. The various sensito- 
metric constants tabulated and the specification of conditions under 
which the determinations were made are as follows: 

All of the sensitometric strips were developed in a solution made up 
according to the two-solution pyro formula given in the section on 
development* used at a temperature of 20 C. 

F. The particular value of fog given is that obtained for the 
development time giving 7 = 1.0 

Td(y = 1-0). This is the development time in minutes required 
to give a 7 = 1.0. 

700. The value of gamma infinity shown in the table is that 
determined experimentally by extrapolation of the time-gamma curve. 
In the construction of this curve, development time was sufficiently 
prolonged so that the curve showed a definite tendency to become 
parallel to the D-log E axis, thus decreasing the amount of extrapola- 
tion required to obtain a fair estimate of the value of 7,,, . 

K. The values for this term as shown in the table were computed 
from the equation dy/dt = K(y m y). By determining graphi- 
cally the slope of the time-gamma curve at the point where 7 = 1.0, 
the value of dy/dt at that point was obtained and when substituted in 
the equation above, together with the already experimentally de- 
termined value of 7 a, , permits the computation of K. The values 
shown in the table, therefore, may be termed the instantaneous value 
of K for the condition 7 = 1.0, and consequently for the time of 
development as shown in the fourth column. Since many photo- 
graphic materials do not conform to the classical H & D theory 
with respect to growth of gamma with increasing development times, 
it follows that K cannot be regarded strictly as an invariant constant 
for all materials. It seems, therefore, that from the practical stand- 
point a value of K, as determined above, may be of somewhat greater 
value than one computed as previously described by using two ZMog 
E curves obtained for development times of T and 2T, respectively. 

* J. Soc. Mot. Pict. Eng., XVH (November, 1931), No. 5, p. 700. 



88 LOYD A. JONES [J. $. M. p. E. 

Obviously it is, of course, possible to put the value of K and 7^ 
shown in the table back into the above equation and compute dy/dt. 
The value of this term, as has already been pointed out, is useful in 
obtaining some idea as to the processing latitude of the material, that 
is, the variation in the time of development which is permissible for a 
specified gamma tolerance. 

L. The values of latitude shown are those given by the material 
when developed to a contrast at which the material in question 
is customarily used. This conveys definite information as to the log 
exposure scale which can be rendered with non-linear distortion. 

i. Values of inertia are those of exposure at the point where the 
straight portion of the characteristic curve having a gamma value of 
1.0 cuts the log E axis. They are expressed, of course, in terms of 
visual candle meter seconds, m.c.s., of radiation of daylight quality. 

10/i. Sensitivity values given in the last column of the table are 
obtained by using 10 as a constant in the expression for speed 



The constants as shown in Table XV are probably as useful as any 
for the purpose of specifying numerically the characteristics of a 
photographic material. Emphasis should again be given to the 
statement already made to the effect that it is hopeless to attempt to 
convey in any set of numerical values as much information as can be 
deduced from a complete set of graphical representations of the sensi- 
tometric data. While numerical constants are very convenient and 
useful for many purposes, they should not be expected to serve as a 
substitute for the more comprehensive graphic representation. 

From a consideration of what has already been said relative to 
the interpretation of sensitometric data, it should be evident that it is 
quite impossible to formulate any single interpretative method which 
will meet the requirements of all of the purposes for which sensito- 
metric data may be required. Interpretation must depend to a great 
extent upon the use for which the information is intended. For 
purposes of standardization sensitometry, it may be desirable to adopt 
a single developer in which all materials to be compared are developed, 
and to express a group of numerical constants derived in such a 
manner as to facilitate intercomparison between the various materials. 
For the control of uniformity of product an entirely different procedure 
may be necessary. For instance, it may be necessary to adopt a 
particular developing solution and technic for each different material, 



Jan., 1932] PHOTOGRAPHIC SENSITOMETRY 89 

and to maintain this with high precision over long periods of time. 
Emphasis may be laid upon the determination with utmost precision 
of some particular characteristic, such as speed or contrast for fixed 
time of development. From the standpoint of the user of a photo- 
graphic material it may be necessary to establish a sensitometric 
procedure which duplicates precisely the processing conditions which 
exist in practice, and lay particular stress on the maintenance of this 
equality at the expense of other factors. If, for instance, sensito- 
metric data are to be used for the control of the uniformity of the 
product turned out by continuous developing machines in the motion 
picture laboratory, great care must be taken to insure that the sensito- 
metric strips are developed under conditions identical to those occur- 
ring in the developing machine. If such is not feasible, it is necessary 
to establish, by an extended series of experiments, a correlatoin 
between the results obtained by some adopted sensitometric condi- 
tions and those existing in practice. It is quite possible that it may be 
necessary to develop particular methods for analyzing the data. For 
instance, in the sensitometric work done in connection with the 
photographic reproduction of sound it is frequently more convenient 
to plot transmission of the silver deposit as a function of log exposure. 
In such cases, of course, it is only necessary to transform density to 
transmission and plot this as a function of log exposure. The analyses 
of these curves require special treatment. It may be useful for some 
purposes to express transmission as a function of exposure rather than 
of log exposure. There are almost numberless variations in inter- 
pretative methods. It is quite impossible to treat all of these 
completely at this time, but it is hoped that the subject matter which 
has been presented may form a foundation upon which further 
elaboration of analysis and interpretation may be built. 
( Concluded in the March issue of the JOURNAL) 
REFEBENCES 

100 NIETZ, A. H.: "Theory of Development," Monograph No. 2 from the 
Kodak Research Laboratories, Eastman Kodak Co. (1922). 

101 SHEPPARD, S. E.: Phot. /., 50 (n. s. 66) (1926), p. 190. 

102 LUTHER, R.: Trans. Faraday Soc., 19 (1923), p. 340. 

103 JONES, L. A., AND RUSSELL, M. E.: Proc. Seventh Intern. Congress Phot., 
(1928), p. 130. 

104 JONES, L. A.: /. -Frank. Inst., 202 (1926), pp. 177, 469, 589; 203 (1927), 
p. Ill; 204 (1927), p. 41. 

106 MEES, C. E. K., AND SHEPPARD, S. E.: "Investigations on the Theory of 
the Photographic Process," Longmans, Green & Co., London (1907). 



THERMIONIC TUBE CONTROL OF THEATER LIGHTING' 

BURT S. BURKE** 



Summary. The development of thermionic tubes has opened an entirely new 
field in control of theater lighting. This development has made possible the ob- 
taining of preset dimming proportional dimming, and a small compact switch- 
board such as has been heretofore impossible. 

The preset dimming feature allows an 'operation whereby a board may be set up 
for any desired number of effects in advance, so that these effects may be called for 
at the will of the operator by operating a single control. This feature might be termed 
an ability of the switchboard to learn effects and bring them out when called upon 
by its master, the operator. 

Proportional dimming, a new feature, allows the lights to be controlled in such a 
manner that they may be dimmed out in combinations while retaining the same color 
tone throughout the dimming process. 

The third desirable feature is that a small compact control board may be so arranged 
that it can be placed as desired in the orchestra pit, or some similar location so that 
the operator becomes a light artist, taking his place in the performance along with 
the organist or other artists. 

In the past 15 years the equipment for controlling illumination in 
theaters has developed from the simple knife-switch type of switch- 
boards to the complex arrangement of circuits and dimmers which are 
required by the elaborate stage productions of the present day. Two 
general qualities are essential in a modern theater switchboard: 
flexibility in the selection of circuits, and flexibility in controlling the 
light intensity of these circuits. The first has been well provided for 
in the various types of multi-preset switchboards which have been 
built for the past several years. The second requirement, as provided 
for in the dimmer systems which are built into the usual multi-preset 
switchboard, leaves much to be desired. 

In the past three years a new means for controlling the intensity of 
the light circuit has become available. The development of thermi- 
onic devices for industrial uses has made possible new systems for 
accomplishing the complex lighting effects required in modern 

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. 
** Westinghouse Electric & Manufacturing Co., East Pittsburgh, Pa. 
90 



CONTROL OF THEATER LIGHTING 91 

theaters. Architects, illuminating engineers, and theatrical producers 
are now able to use lighting in ways that have never before been 
possible. 

The thermionic tubes used in theater lighting equipment have been 
of the hot cathode grid-glow type, whose output may be regulated by 
properly controlling the grid circuit of the tube. The whole control 
system consists briefly of a reactance type dimmer, similar to the 
reactance dimmers which have commonly been in use for many years, 
a thermionic tube unit for supplying direct current to saturate the 
reactance dimmer and a control means consisting of a network of 
potentiometers for properly controlling the output of the tube unit. 

A system of this type was used for the control of the stage and 
auditorium lighting of the new Los Angeles theater which was opened 
January 30, 1931. This paper will describe this equipment, which is 
typical of the thermoionic tube type of control for theater lighting. 

The Los Angeles theater, similar to many theaters of its size, is 
equipped for both motion pictures and stage presentations. On this 
basis, the control switchboard was split into two parts, one of which 
was located in the projection room for controlling the auditorium 
lights, and the second of which was located at the stage floor for 
controlling the stage lights. On the stage floor is also located a 
remote panel which allows the operator to obtain color master control 
of the auditorium circuits as well as control of the five-scene preset 
arrangement which will be desired later. Dual control for the foot- 
lights, the first border, the orchestra floods, and the stage floods is 
provided so that these circuits may be controlled either from the stage 
floor or from the projection room. A transfer scheme is used to 
transfer these controls from one place to the other so that no inter- 
ference of control is possible. 

Each of these two switchboards consists essentially of a reactance 
dimmer bank, a set of tube units for use in connection with the d-c. 
coils of these reactors, and a control board used to control the 
output of the tube unit, thus indirectly controlling the intensity 
of the lighting circuit. 

The reactance dimmers are similar to the standard theater duty 
reactance dimmer in that they consist of a-c. coils and a d-c. coil 
mounted on an iron core. The reactance of the dimmer is varied by 
changing the amount of direct current in the d-c. leg of the reactor, 
thus varying the saturation of the iron. With no direct current in 
this coil, the iron is unsaturated and of very high reactance, thus 



92 BURT S. BURKE [j. s. M. p. E. 

dimming out the light circuit with which it is associated. By increas- 
ing the direct current the iron becomes saturated, so that the reac- 
tance of the dimmer becomes lower, thus increasing the voltage across 
the lamps and bringing them up to full brilliancy. 

The direct current supplied to the coil of the reactance dimmer 
comes from the thermionic tube unit which consists of two grid-glow 
tubes, a control tube, and the requisite transformers for supplying 
the proper filament and plate voltages for these tubes. The tube 
unit receives its power from the 115-volt, 60-cycle, single phase mains, 
which is applied to the two grid-glow tubes through a suitable trans- 
former in order to obtain the proper voltage on the plates of the tubes. 
The a-c. voltage is rectified by these tubes and the resulting rectified 
alternating current is impressed on the coil of the reactor. The 
magnitude of the d-c. output of the tube unit is varied by changing 
the relation of the grid voltage to the plate voltage of the grid-glow 
tubes. This is done by means of the control tube, which is a vacuum 
tube similar to the standard UX-226 tube used in radio circuits. The 
output of this control tube is varied by a system of potentiometers 
located on the control board which will be described in the following 
paragraphs. One set of these tube units was mounted on the reactor 
bank located in the projection room for the auditorium lights, and a 
second set of the tube units was mounted on the reactor rack located 
in the basement which was for the stage lights. 

The two control switchboards, one located in the projection room 
for the auditorium circuits, and the second located on the stage floor 
for the stage circuit, are similar except for the circuits controlled, and 
may be briefly described as follows: 

For the control of each individual circuit there is provided a pilot 
switch, a selector switch, an indicating lamp, 5 preset potentiometers, 
and an individual control potentiometer. In addition to this there 
is provided for each circuit a scene-fader by means of which it is 
possible to fade from one effect into the next effect giving a gradual 
transition from one to the other. These faders are ganged together 
and have a common drive, operated either by a handwheel or a 
motor. 

The pilot switch serves a purpose similar to that of the pilot switch 
on the standard switchboard. That is, one throw of the pilot switch 
connects the circuit directly to a hot-bus connection so that the circuit 
may be controlled independently of any of the master set-ups. A 
second position of this switch connects the circuit through a color 



Jan., 1932] CONTROL OF THEATER LIGHTING 93 

master control so that it is possible to control an entire color by 
means of a single color master. The middle position of the switch is 
the off position. The pilot lamp is wired in connection with the 
pilot switch so as to indicate when the circuit is hot; that is, when 
the switch is thrown directly to the hot bus position or when the 
pilot switch is thrown to the color master position and the color 
master is energized. 

The selector switch is used for transferring the control from the 
individual control potentiometer to the preset potentiometer. In 
one position of this switch the grid lead from the tube circuit is 
connected to the moving arm of the individual potentiometer. In 
this position, the output of the tube circuit may be controlled by 
manipulating the individual potentiometer, and it is not affected 
by any changes made in the preset potentiometer. In the second 
position of the selector switch the grid lead is connected through the 
fader to the preset potentiometer. In this position the circuit is 
controlled through the preset control by means of which the intensi- 
ties of the circuit may be set for five scenes in advance. 

Regarding the operation of the preset potentiometer, let us first go 
back to the operation of the tube unit. As has been previously 
stated, the output of the tube unit, and consequently the intensity of 
the lighting circuit, is changed by altering the grid potential on the 
control tube. The grid potential is obtained through a system of 
potentiometers from the d-c. control source as indicated on the 
attached diagram. (Fig. 1.) 

This preset operation is accomplished by two methods, one of 
which allows a gradual fading from one effect to the next and a second 
of which permits the operator to flash immediately from the effect 
in progress to any other effect that has been previously set up. These 
two operations are accomplished by means of the dimming fader and 
the multi-contact flashing relays as described in the following para- 
graphs. 

One dimming fader is provided for each circuit. In addition to 
this there are five multi-contact flashing relays, 1 to 5, and a fader 
disconnecting relay (/). (Fig. 1.) One contact of each relay is con- 
nected in each control circuit as shown on the diagram. The normal 
position of the relays for operating through the dimming fader is to 
have relays 1 to 5 open and relay (/) closed. Now to set up circuit 
No. 1 for full intensity the slider of preset potentiometer No. 1 is 
moved to the positive end of the potentiometer. To set up circuit 



94 



BURT S. BURKE 



[J. S. M. P. E. 



No. 2 for black-out the slider is moved to the negative end. In 
order to obtain intermediate intensities on the other scenes, the 
sliders are moved to intermediate positions corresponding to the 
intensities desired. 

Now assuming that the switchboard is operating on scene No. 1, 
the pointer of the dimming fader will be connected to point 1 on 
this piece of apparatus. In order to transfer to scene No. 2, the 
dimming fader, which consists of a unit for each circuit to be con- 
trolled on a common drive is moved either by a handwheel or by 
means of a motor drive to position No. 2. It can be seen, therefore, 




!> LW 

H h h h h 





MULTI CONTACT #LArS 



FIG. 1. 



Schematic diagram of scene flashing control thermionic type 
theater switchboard. 



that since point No. 1 of the dimming fader is connected to the slider 
of preset potentiometer No. 1 and is therefore at the same potential 
as its preset for this scene, and since point No. 2 is connected to the 
slider of preset potentiometer No. 2, there is a gradual transition from 
the potential set for preset No. 1 to that for preset No. 2. This gives 
a gradual change on the potential impressed on the grid circuit of the 
tube unit, and, therefore, a proportional change in the lighting 
circuit. 

A similar operation is also performed to transfer from scene No. 2 
to scene No. 3. It is, furthermore, possible with this type of equip- 



Jan., 1832] CONTROL OF THEATER LIGHTING 95 

ment to set up the scenes not in use for additional effects without 
affecting the scene in progress. 

The purpose of the scene flashing equipment is to allow the operator 
to transfer immediately from the effect that may be set up from one 
scene to the effect set up for any other scene and have all the lighting 
circuits come to the desired preset intensity. This is accomplished 
by means of the relays, 1, 2, 3, 4, 5 and / which operate to disconnect 
the grid lead from the scene fader and connect it to the preset 
potentiometer associated with the relay selected. Thus the potential 
which has already been set up on this potentiometer is applied to the 
grid lead of the tube unit and a corresponding intensity of the lighting 
circuit results. When any other relay is operated, the relay previously 
closed is automatically disconnected by means of the switch keys, 
which are interlocked. 

In order to obtain color master operation of any of the circuits, as 
has been previously described, the pilot switch is thrown to the color 
master position. This transfers the lead to the positive end of the 
control potentiometer from the positive bus and connects it to the 
sliding arm of the color master potentiometer. Thus it may be seen 
that the voltage on all the potentiometers connected to this particular 
color master is varied by moving its sliding arm. Consequently, a 
proportional change in the voltage impressed on the sliding arm of the 
individual potentiometers is obtained, which results in a proportional 
change in the lighting intensity of the circuits connected to these 
controls. Thus, if one of the circuits connected to this color master is 
at full brilliancy, a second at 3 /4 brilliancy, a third at l / s brilliancy, 
etc., should these circuits be dimmed out by the color master they 
would start dimming at the same time, and proportionally change, so 
that they would reach the black-out position at the same time. This 
is in contrast to the operation of the standard interlock type of color 
master in which a similar operation would result in the circuit at full 
brilliancy being dimmed until it corresponded to the one at */4 
brilliancy, at which point the second circuit would interlock with the 
color master and both these circuits would travel until they reached 
half brilliancy, where the third circuit would interlock and finally 
all would black-out together. This often results in a spotty effect and 
is undesirable. By using an electrical color master rather than a 
mechanical color master, a proportional dimming effect is accom- 
plished as previously described. 

In order to obtain grand master control, the pilot switch on the 



96 



BURT S. BURKE 



[J. S. M. P. E. 



color master section is transferred from the hot-bus position to the 
grand master position which connects the circuits on the color master 
to a master generator. By varying the voltage on this master 




FIG. 2. Thermionic lighting control console, Severance Hall, 
Cleveland, Ohio. 

generator, the potential impressed on the color master is varied, thus 
causing a proportional change similar to that previously described for 
the circuits connected to the grand master. 



Jan., 1932] CONTROL OF THEATER LIGHTING 97 

Motor operation is provided for the dimming fader in the Los 
Angeles theater so that it is possible for the operator to change from one 
effect to another simply by pushing a small telephone switch starting 
the motor drive. This motor drive is so provided with limit switches 
that it will travel to the succeeding scene at which point it will stop, 
and will not start again until the operator pushes the "start" button. 




FIG. 3. 5-Scene thermionic control board for stage 
lighting control, Los Angeles Theater, Los Angeles, Cali- 
fornia. 

This allows an easy means of obtaining remote control of the intensity 
of all the lighting circuits in the theater. In the old type of multi- 
preset board it was possible to obtain remote control scene changes, 
but it was impossible to preset the intensities, as it was necessary for 
the operator to set the dimmers beforehand or to change their setting 
when transferring from one effect to the other. By using the thermi- 



98 BURT S. BURKE [J. S. M. P. E. 

onic tube control, it is possible to preset the dimming as well as to 
preset the circuits which are to be used and, furthermore, to obtain 
remote control of these circuits if it is desired. This is of special 
advantage in motion picture houses where it is desirable to control 
the light from the projection room and at the same time have a switch- 
board on the stage floor which can be used in case of stage presentation 
work. 

In the Los Angeles theater, the remote control board at the stage 
floor allows the stage switchboard operator to have full control of the 
color masters for the auditorium circuit, and in addition to this allows 
him to change the lighting effects on the auditorium for five presets 
which had previously been determined by means of the switchboard 
located in the projection room. Furthermore, by means of the color 
masters, it is possible for the operator to dim out any particular color 
from any scene that had previously been preset, thus giving a very 
flexible control. 

The previous description is typical of this type of control. Due 
to the rapid development in the art, at least one other scheme has 
already been conceived. Instead of a grid-glow type tube, a vacuum 
tube of rather large plate capacity is used and a motor generator set 
supplies 500 volts d-c. to the plates. The output is regulated by grid 
control of these tubes. That is, the 500-volt supply from the genera- 
tor is connected in series with the vacuum tube and the d-c. coil of 
the reactor. Thus, by varying the impedance of the vacuum tube by 
change of grid potential, the amount of direct current that is allowed 
to pass through the d-c. coil of the reactor is changed. The control 
equipment, that is, the potentiometer set-up is practically a duplicate 
of that used in the Los Angeles theater, the new developments 
affecting the tube units rather than the control. 

Another advance has been in the method of control by means of 
which it is now possible for a switchboard to be built wherein the 
operator may change from one scene to any other scene and get a 
gradual fading effect from the one to the other, or to obtain a flashing 
effect as previously described. The Los Angeles theater was built 
prior to this development, so that it was necessary to fade from one 
scene to the next succeeding scene. The new development allows a 
much more flexible control due to the fact that in stage presentation 
work it is often desirable to repeat an effect ; with this type of control 
it is possible to do this as often as is desired, and to fade into this 
effect from any other that may be in progress. 



Jan., 1932] CONTROL OF THEATER LIGHTING 99 

Another interesting development that has recently been brought 
forth, the first application of which is for the control lighting of the 
Buckingham Fountain in Chicago, provides a continuous preset pro- 




FIG. 4. Reactor and thermionic unit rack, Los Angeles Theater, 
Los Angeles, California. 

gram which is laid out in advance for an evening's performance. 
This is accomplished by means of an insulating track on which there 
has been drawn a conducting path. This moves and is continuously in 



100 BURT S. BURKE 

contact with a potentiometer similar to the control potentiometers 
for the tube units. By varying the location of the conducting strip 
on the moving track, a variable preset potential is obtained which 
correspondingly changes the output of the tube unit, and a change 
in the intensity of the lighting circuit results. While this is developed 
for floodlighting control it may be used for such an application as 
varying the lighting of a theater according to a definite program for 
the overture. It could also be used for providing a light change 
program for the patrons at the time between the opening of the 
house and the beginning of the performance. This program is motor 
driven, and can be started and allowed to run for a definite period of 
time after which, by means of a transfer relay, the control can be trans- 
ferred back to the regular stage switchboard for use in connection with 
the picture or the stage presentation. 

In summarizing it may be said that the application of thermionic 
tubes to theater dimming has made possible a stage board giving the 
following very desirable features: 

(1) Presetting of intensity for any desired number of scenes. 

(2) Proportional dimming. 

(3) A light compact board using telephone switches, thus insuring ease of 

operation with added assurance of proved reliability of this type of equip- 
ment. 

(4) Low control voltage (less than 50 volts, d-c.) allows use of telephone 

cable for control wiring. 

(5) Remote control easily added. 

In fact, the field of application of tube control to lighting is in its 
very infancy, and due to the rapid development in tubes, as we have 
witnessed in the past in the radio field, a great deal may be expected of 
this type of equipment. 



A PORTABLE NON-INTERMITTENT CINE PROJECTOR* 

Summary. A portable projector made by the tablissement Gaumont Franco- 
Film Aubert is described. The projector is of very small weight and is arranged 
for carrying in a case. The film moves with a constant motion past the axis of the 
light source and the projection lens, the image being maintained stationary upon 
the screen by a combination of the movement with an optical "compensator." It 
is claimed that due to these features, wearing of the film has been very much reduced 
and the motion is extremely silent in operation. The article describes briefly the 
optical principle of the motion, how the principle is applied, and the construction 
and assembly of the apparatus. 

The "Simplicine"" is a cin6 projector for standard film, self-con- 
tained and complete, yet small enough in bulk and weight to be port- 
able. The whole projector is enclosed in a metal casing and can be 
carried easily on a sling strap. Its erection is almost instantaneous 
and its manipulation so simple that no special experience is required 
for its use. 

The chief importance of this machine, particularly so far as the 
non-professional user is concerned, is that it employs the principle of 
constant movement projection. The film moves with a uniform mo- 
tion across the axis of the light source and the projection lens. This 
is a vital difference from the usual intermittent projector, in which a 
Maltese cross or other mechanism is used to drag the film into position 
and then bring it momentarily to a standstill in the gate of the 
machine. In the new projector the image is kept stationary on the 
screen by means of a special combination of the movement with an 
optical device termed a "compensator." One greater advantage of 
such a system is the very much reduced wear on the film perforations 
owing to the elimination of the violent and repeated tugs to which 
films are subjected in ordinary types of projectors. Film is said to 
last five times as long when it is run in this continuous manner. To 
this advantage may be added the not less important one that abso- 
lutely silent mechanism can be obtained when all the moving parts are 
given nothing but continuous rotary movement, as is the case in the 
"SimplicineV' 

* Translated from Revue d'Optique, 10 (April, 1931). No. 4, p. 178. 

101 



102 



NON-INTERMITTENT PROJECTOR 



[J. S. M. P. E. 



The Optical Principle. Fig. 1 represents a film moving in a down- 
ward direction and carrying a series of images 1, 2, 3, etc. Imagine 
that in front of these images is a series of similar lenses Oi, 2 , 3 , etc., 
each having its focal point in the plane of one of the images, and 
suppose this chain of lenses to move in a direction parallel with the 
film and at the same speed. If the beams of parallel light so formed 
meet a fixed lens C the images of the different elements of the film 
will be superimposed in the focal plane of this lens. Indeed, if we 
consider an element formed by an image on the film and the corre- 
sponding lens, the image of a point of this element given by C will 
have its position, in the focal plane of C, determined solely by a 
straight line passing through the optical center of C and parallel to 
the straight line joining the given point in the element to the optical 
center of the corresponding lens. Now this straight line as it moves 
remains parallel, consequently the final image is fixed. This is true 



o,t 

*& 



' C 



FIG. 1. Diagram illustrating the optical principle. 

for all images of the points of the element of the film. The straight 
lines joining corresponding points of the element to the optical centers 
of the corresponding lens being parallel, the images of successive 
elements are superimposed in the focal plane of C. Hence the pro- 
jection screen E is made to take the position of the focal plane of C 
and focusing for various distances is obtained by providing a set of 
lenses C of different focal lengths. 

How the Principle Is Applied. The realization of this principle in 
actual fact has been achieved in the following manner: The lenses are 
set round the periphery of a cylindrical drum T (Fig. 2), which is 
free to turn on its axis. T is made to rotate by the fact that the film 
catches a tooth D formed on the drum and carries T round with its 
own movement. The lenses therefore move at the same speed as the 
film. Light passing through the illuminated film reaches the lenses 
O after traversing a prism P (Fig. 2), which is formed integrally 



Jan., 1932] 



NON-INTERMITTENT PROJECTOR 



103 



with the fixed axis of the drum. This prism has two reflecting sur- 
faces MI and Mz set perpendicular to one another. The system thus 
produced is that indicated in Fig. 1, with the difference that the film 
and the lenses do not travel in a straight path but follow curves 
of the same radius. This difference has a practically negligible effect 
on the quality of the images, assuming that the film is illuminated 
only over the length of two images. This means that two images 
and two only of the chain of lenses are actually utilized, film elements 
and lenses which have any appreciable inclination to the normal 
being kept out of action. Furthermore, the projected image shows 




FIG. 2. The construction of the drum 
carrying the lenses on its periphery. 

its maximum illumination at the moment at which the corresponding 
lens has its axis coincident with the axis of projection, and the effect 
of this is to reduce greatly the aberrations of the images thrown by 
adjacent lenses which are slightly inclined. This mechanism is 
exceedingly simple: it consists of a single component moving with a 
continuous rotary movement at low speed 80 revolutions a minute 
for a projection speed of 16 pictures a second. Wear is therefore 
reduced to a minimum and the running is quite noiseless. 

The Projector Described The "Simplicine*" has been given the 
form of a rectangular case, the top and side of which are formed with 



104 NON-INTERMITTENT PROJECTOR [J. S. M. P. E. 

hinged swinging sections which are raised vertically for use. All the 
mechanism is then made visible. 

The feed-reel, 3 (Figs. 4 and 5), with its pulley, 4, is then fixed on 
this raised portion of the casing. The film passes under the feed 
sprockets, 5, and on to the drum T carrying the compensating lenses ; 
then on to the toothed sprocket 6, the take-up reel 7, rotated by its 
driving pulley 8. 

Masking the film on the screen is effected thus: when the lever, 
9 is pressed downward, the roller, 10 is pushed up between the two 
pressure rollers, thus raising the film and raising the roller, 11. If 
the pressure on this lever is released, the roller 11 returns into contact 





FIG. 3. External view of projector. 



FIG. 4. View of projector opened 
for use, showing internal arrange- 
ment. 

with the drum, and the loop formed by the film is taken up by the 
movement of the drum. The importance of the formation of this 
loop is that in this operation the film advances by one perforation. 
Masking is thus effected by the displacement of the image by an 
amount equal to a quarter of its height. 

Motor Drive and Lighting. The driving parts are carried on a 
fixed aluminum platform and can be removed as one unit from the 
box. This block consists of an electric motor, 12; a pulley drive, 13, 
reversible for rewinding; and a transformer, 14, to feed the lamp 15. 
In front of this assembly, against one of the panels of the box, are 
mounted side by side two rheostats, for the lamp and for the motor, 



Jan., 1932] 



NON-INTERMITTENT PROJECTOR 



105 



with finger controls, 16 and 17, for their adjustment; and an am- 
meter, 19, for the control of the lighting system. A plug let into the 
casing provides for the connection of the apparatus to a source of 
electric power. 

The lamp used is a 225-watt Phillips, taking 30 amperes, at 7 1 A 
volts. The beam is of about 525 cp. in a horizontal direction per- 




FIG. 5. Diagram of internal arrangement. 

pendicular to the tungsten filament. The lamp is conical in shape 
and works upside down in order to avoid the blackening of the bulb 
around the filament. The lamp is suitably supported and adjusted. 
The optical system includes a two-lens condenser 25, and spherical 
mirror 26 in line with the axis of the filament. 

As a safety device there is a wire gauze, 27, which is arranged to 



106 



NON-INTERMITTENT PROJECTOR 



come into place automatically between the light source and the 
condenser when the film is stationary. This takes place by a centrif- 
ugal action. Its effect is to protect the film from any dangerous 
degree of heating without restricting the light unduly when the 
machine is used for still projection from selected pictures. 

As regards focusing, the apparatus possesses five collimating 
lenses C (Fig. 1), arranged on a rotatable disk, and by means of these 
the image can be focused on a screen at any distance from 6 to 32 
feet. The milled edge of this disk projects through the casing at the 
side so that it can be rotated by the finger, and a spring detent sets 
it in accurately centered position whichever lens is in action. A 




FIG. 6. Showing the method of 
rewinding. 

rectangular window in the front wall of the apparatus, made to allow 
the beam to pass, is fitted with two sliding covers adjustable verti- 
cally to cut off parasitic images which would otherwise be thrown on 
the screen. 

Re-winding at the end of the projection is very simply carried out. 
The film is released from the drum and from the guiding sprockets, so 
that it runs as shown in Fig. 6. An adjustment is then made, to 
allow the take-up reel to turn freely on its axis and to fix the feed- 
reel to its axis. The full take-up reel thus becomes the feed-reel, 
and vice versa. The motor, running just as in projection, then 
rapidly re-winds the film, leaving it ready to project again. 



COMMITTEE ACTIVITIES 

REPORT OF THE PROJECTION PRACTICE COMMITTEE* 

The Projection Practice Committee wishes to direct attention to 
what it considers one of the foremost causes of waste and monetary 
loss suffered by the motion picture industry, namely, the mutilation 
of positive prints. This mutilation not only results in a consider- 
ably shortened life of the individual print, which is serious enough 
in itself, but in addition to this, it is impossible to obtain the optimum 
screen results, which are so highly important in creating the proper 
illusion so necessary to the motion picture play. Both picture and 
sound are affected by mutilation of film. 

It is generally understood that the mutilation of film is frequently 
due to the maladjustment of projector parts, wearing of projector 
parts, accumulation of emulsion during projection, excessive oiling 
of projector or leakage of oil, and careless handling of film. The 
Projection Practice Committee is of the opinion that there is urgent 
need for the establishment of standards dealing with the various 
tensions to which the film should be subjected during projection, 
the clearances of adjacent projector parts and sound apparatus, 
allowable tolerances, and the amount of wear projector parts may suffer 
without impairing the quality of the picture or causing mutilation of 
film. 

The committee, therefore, plans to conduct a thorough investigation 
which will be nation wide in scope, with the view of obtaining all 
necessary data for submittance to the Society for the purpose of 
adopting such standards. In order to accomplish this, the committee 
requests the earnest cooperation and support of the Society as a 
whole, as well as of associated individuals and organizations. Their 
assistance will be needed as this work will be of considerable mag- 
nitude and should, when completed, prove invaluable to the industry. 

The Committee wishes also to call attention at this time to the lack 
of uniformity in the processing of prints, which constitutes another 

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. 

107 



108 PROJECTION PRACTICE COMMITTEE [J. S. M. p. E. 

serious loss. In regard to the processing of film, there seems to be no 
standard for this work at the present time. One producer uses a 
certain method of processing film; another producer simply waxes 
the margins of the print; and a third producer does not process 
the print at all. This condition works a hardship on all concerned, 
inasmuch as it frequently happens that the producer who has 
processed his product suffers by reason of the fact that the theater 
uses unprocessed film at the same time. This evil adversely affects 
both the sound quality and the quality of the picture. 

It is well known that with unprocessed film there is a tendency to 
accumulate emulsion at the tension points in the projector. Forma- 
tion of emulsion greatly increases the tension applied to the film 
and imposes a strain on the sprocket holes. Occasionally a positive 
print is irreparably damaged during its first projection. The Pro- 
jection Practice Committee recommends that a thorough investiga- 
tion to find the best method or methods of processing film be con- 
ducted by a designated committee of the Society so that such methods 
may be recommended as a standard for the industry. 

Unless such a standard is adopted, generally accepted, and put into 
use by the producers of film, the industry will continue to suffer 
the great loss now occasioned through faulty (or the lack of) process- 
ing methods, and such benefits which should accrue through the 
adoption of the standards relating to projector tensions, adjust- 
ments, etc., would be largely nullified. In the opinion of the Pro- 
jection Practice Committee, such a work is one of the most impor- 
tant contributions the Society could make to the industry. 

RESOLUTION 

The Projection Practice Committee wishes to include in the records of the 
Society a statement of its appreciation of the splendid work and cooperation 
which President Crabtree extended to this Committee and, also, for his realiza- 
tion of the important role which practical projection plays hi the motion picture 
industry. 

Through President Crabtree's foresight, initiative, and efforts, a committee 
to deal with the practical problems of projection was formed for the first time 
in the history of the industry, and specific problems greatly in need of attention 
and correction were brought to the light of day and taken under consideration. 

Therefore, we, the Projection Practice Committee, gratefully acknowledge 
what President Crabtree has done for the craft, for the Society, and for the 
industry at large, and extend to him our thanks and a vote of confidence in his 
conduct of the affairs of the Society. 



Jan., 1932] PROJECTION PRACTICE COMMITTEE 109 

HARRY RUBIN, Chairman 

THAD. C. BARROWS R. H. McCuixouGH 

G. C. EDWARDS P. A. McGuiRE 

SAM GLAUBER RUDOLPH MIEHLING 

J. H. GOLDBERG F. H. RICHARDSON 

CHAUNCEY GREENE MAX RUBEN 

HERBERT GRIFFIN H. B. SANTEE 

JESSE J. HOPKINS L. M. TOWNSEND 

DISCUSSION 

MR. McGuiRE: For quite a few years I was one of those who vigorously 
protested against the neglect of projection by this Society, but no longer have 
I any cause for complaint as we have our own Projection Practice Committee 
and it is up to ourselves to make good. In discussing the Report of the Pro- 
jection Practice Committee it is not my intention to complain or criticize, but 
to offer some suggestions which I hope will be helpful to the motion picture 
industry and a benefit to the Society of Motion Picture Engineers. I ask you 
to be patient because much of what I will say is to be of a somewhat general 
nature and, perhaps out of place in the proceedings of this Society. But most 
of the papers and much of the discussion of the Society are more or less incompre- 
hensible or relatively unimportant to some part of the membership of this organi- 
zation. In order to deal only with subjects which would be of interest to every- 
one it might be necessary to hold a hundred conventions or divide the meetings 
into an equally large number of groups. In its proceedings, the Society of 
Motion Picture Engineers must give some attention to invention, development, 
manufacture, maintenance and operation, the electrical, chemical, and mechanical 
divisions of the industry, visual and sound recording and reproducing, and 
always theory and practice. These are rough classifications, but give a general 
idea of the vast field the Society must cover. 

The Society of Motion Picture Engineers is not a scientific body seeking ab- 
stract truth, but a technical organization with a very definite commercial back- 
ground. When we lose sight of the fact that we are part of the motion picture 
industry we fail to realize the true purpose of the Society. It, therefore, seems 
to me that anything the Society can do to render a practical service to the in- 
dustry should result in the organization receiving increased support. The 
benefits that the industry has derived from the Society of Motion Picture Engi- 
neers have not always been recognized because they were often of an extremely 
indirect and intangible nature. The fact that the Society has for many years 
focused attention upon the technical side of the motion picture industry and to 
some extent has won the interest of non-technically minded executives is in 
itself a great achievement. The executives of this industry have never given 
the Society adequate support, and I believe that the producers and exhibitors 
have contributed more to a single activity of another organization in this 
field than they have to the Society of Motion Picture Engineers in its entire 
history. 

The Society is facing new conditions and it is desirable that the service which 
it renders the industry should be more direct and more obvious. If this can be 



110 PROJECTION PRACTICE COMMITTEE [J. S. M. P. E. 

done, the Society will receive increased support and be in a better position to 
carry on its important work. This organization is in a particularly strong 
position to secure technical data regarding the cause and prevention of film 
mutilation. Various attempts have been made to get this information, but there 
is good reason to believe that the results have not been entirely satisfactory. 

Someone has said that "science is common sense made exact." The Projection 
Practice Committee will conduct a scientific survey, collecting the facts sys- 
tematically and thoroughly, and present them in an authoritative report. 
When this is done definite action should result and the Society will have rendered 
a service comprehensible in terms of dollars and cents. 

The work we are undertaking, however, will involve considerable time and 
expense, and should receive adequate support from the Society as well as the 
industry. It is an unfortunate fact that the industry does not take proper 
interest in the collective thought developed by such an organization as the 
Society of Motion Picture Engineers. Progressive projectionists in this organi- 
zation, and their own projection societies, are constantly giving their own time 
to do valuable technical work without receiving the least recognition from the 
executives of their own firms. Conceding that this is a period in which executives 
are very properly insisting upon economies, it nevertheless seems unwise to 
ignore totally all the collective effort for the betterment of the industry. 

Back of the artistic side of the motion picture industry is a vast technical 
field whose work offers infinite opportunity for flaws and failures. Motion pic- 
tures provide entertainment and education through chemical, mechanical, and 
electrical processes. What the public pays for is not the product of a single 
commercial organization, and it is important that the Society of Motion Picture 
Engineers should bring this to the attention of the industry emphasize the 
interdependence of the various departments and point out the need for coordi- 
nation. In all work which is not of a competitive nature the industry benefits 
tremendously from the collective thought developed in such organizations as 
the Society of Motion Picture Engineers. I sincerely hope that a way will be 
found to encourage and finance adequately the efforts of the Projection Practice 
Committee to find the cause and prevention of film mutilation. The men on 
this Committee have the technical and practical experience to do the work. 
Their report should result in a tremendous saving through prevention of waste 
and the improvement in screen presentation. 

PRESIDENT CRABTREE: I indorse Mr. McGu ire's remarks one hundred per cent. 
Of course the world was not made in a day. But it is encouraging that the pro- 
ducers have shown a much greater willingness to do things for us at this Con- 
vention than at any time previously. 

Will Mr. Griffin give us a few details as to how the stoppages occurred in the 
projection room? 

MR. GRIFFIN: There was only one cause, Mr. President, and that was the 
processing. Emulsion, or whatever was on the film, supposedly to prevent its 
"seizing up" during transit through the mechanism, did not prevent it. In 
some cases it was wax, and in some cases something else. 

PRESIDENT CRABTREE Where did it seize? 

MR. GRIFFIN: In the gates. It can seize anywhere in transit, wherever 
there is tension at the picture gate or the sound gate. This time it happened 



Jan., 1932] PROJECTION PRACTICE COMMITTEE 111 

to be at the sound gate. It is not the fault of the manufacturer of the equipment, 
because one can run a film through that has been run two or three times, and 
properly processed, and it will not cause any trouble at all. We could have run 
the film through without stopping; it was a new print, however, and we wanted 
to save it. And in as much as it was not a very serious matter to stop the picture 
here, as would have been the case in a theater, we stopped it. But the sound 
was terrible in some cases, caused by a piling up of the wax, behind the film, 
thus changing the thickness of the scanning beam. 

PRESIDENT CRABTREE: Was this a new machine? 

MR. GRIFFIN: It was not new in the sense that it had never been run before. 
Films had been run through it on different occasions, but the equipment to all 
intents and purposes is new. 

PRESIDENT CRABTREE: In the case of two metal surfaces, one of which is 
polished to an extremely high degree with rouge, and one which is not polished, 
the polished surface will not pick up as much gelatin or emulsion as the rougher one. 
I was wondering, therefore, that if this machine had been a little older, would 
the trouble have occurred? 

MR. GRIFFIN: The finishes on all parts that come into contact with film 
are finished with rouge, and I believe that RCA uses crocus-cloth for polishing. 
I don't know of anything better. The surfaces are highly polished and burnished, 

MR. SUMNER: I happen to be an exhibitor, and this report of the Committee 
was very interesting to me. We happen to run a theater that is called a "first 
subsequent run;" that is, we run after the key point in this district, which is 
Boston. I have attended a number of the conventions, and have heard the 
reports from the various specialists in the studios; and I realize the great amount 
of thought and work that is put into the pictures, the great mass of work that 
has been done to accomplish perfect sound, and so forth. And yet, when these 
prints get to the theaters, the greater part of that work has been ruined by 
improper handling of film. As an exhibitor, I wish to state that I believe that 
the work that has been begun by this Committee is most important. I want 
to urge them not to stop with the problem of processing film. They must go. 
much further than that. 

In spite of the noiseless recording system, the prints reach the theaters so 
dirty and scratched that the work of noiseless recording has almost gone for naught. 

I think this Committee is one of the most important factors in the organization 
and I want to urge that it be given all possible support in its work. 

PRESIDENT CRABTREE: I should like to ask Mr. Griffin: Was the accumulation 
of emulsion due to friction along the perforations, or at some portion of the 
picture area? In other words, is it necessary to process the entire surface of 
the film, or merely the edges of the perforations? 

MR. GRIFFIN: For projection purposes it is necessary only to process or 
lubricate, as it may be called the edges of the film in the sprocket hole area. 

PRESIDENT CRABTREE: Was the film in question lubricated or "processed?" 

MR. GRIFFIN: I cannot answer that. I do not know either the processes 
or who does the processing. I only know what occurs during projection. 

MR. FAULKNER: Four different prints caused the trouble, and each one 
of the four had four different applications and four different kinds of chemicals 
on them. The gathering of emulsion on three different prints that I looked at 



112 PROJECTION PRACTICE COMMITTEE [j. s. M. P. E. 

was identically in the same spot, showing that no matter what caused it to gather, 
it did so in exactly the same place on the film. I did not see the fourth print 
but the sound quality and the way in which it behaved were similar. 

As Mr. Griffin says, as far as passing the film through a projector is concerned, 
it is only necessary to lubricate the margin of the film. The emulsion that is 
on film, unless the metal parts with which it comes into contact are lubricated, 
is quite likely to stick. Therefore, the film is lubricated for the purpose of 
keeping the tension shoes lubricated. 

Mr. Rubin asked me to present to you his idea that "processing" is an in- 
correct term to use for this process. He wants to find a name for waxing, treat- 
ing, processing or ' 'whatnot, ' ' and to standardize that name. I went to a dictionary 
and ran down every name I could think of. I have a great number of them, 
none of which I think would be appropriate, except perhaps "treatment" or 
"finishing" or the like. "Processing" is used to indicate anything that may 
happen to film from the time it is printed to the time it is developed for screening. 

PRESIDENT CRABTREE: Why not use the word "conditioning?" 

MR. FAULKNER: Some of the names I accumulated are: hardening, com- 
pleting, seasoning, curing, impregnating, finishing, duratizing, dura-proofing, 
inuring, toughening, preserving, protecting, treating, perfecting treatment. 
None of these I think would be satisfactory except perhaps "conditioning" or 
"treating." I do not like "processing," nor does Mr. Rubin. 

MR. McGuiRE: I ask you not to exaggerate the importance of "processing" 
merely because it has received special attention in this discussion. It is a serious 
problem, but we shall have other important subjects to consider in our efforts 
to find the cause and prevention of film mutilation. There has been much talk 
in the past about film mutilation and various organizations have dealt with it 
rather unsuccessfully. 

The Projection Practice Committee is starting out with the idea that there 
seems to be an evil which is called film mutilation, but that it knows absolutely 
nothing about its cause and prevention. We hope to be able to gather some 
data in the next six or twelve months, which will save the motion picture industry 
a tremendous sum every year and greatly improve the quality of screen pres- 
entation. 

MR. J. CRABTREE: I think a little more attention to the projector is what is 
necessary. I often project green film, and find that as long as the projector is 
kept in shape, little trouble is experienced. Mr. Faulkner pointed out that 
last night the accumulation occurred in the same spot in each case, which goes 
to show that there is a high spot somewhere. One cannot expect lubrication 
to take care of all high spots. Eliminate the high spots, and the lubrication 
won't be so necessary. 

MR. GRIFFIN: I must take exception to that. I don't know under what 
conditions Mr. Crabtree projects his prints, but I defy anybody to take a piece 
of green film off the drying rack and project it under conditions existing in the 
theater today and not have it seize up, no matter how well the projector is de- 
signed. 

PRESIDENT CRABTREE: Are you speaking now of a film to the edges of which 
wax has been applied? 

MR. GRIFFIN; Mr. Crabtree said he would use it without treatment right 



Jan., 1932] PROJECTION THEORY COMMITTEE 113 

out of the laboratory. It is not waxed there. Now, waxing is not the solution, 
apparently, because the wax peels off and rolls up. With the old silent machines, 
waxing was all right. Today we have sound. The wax rolls off, gets in the 
sprocket holes, and is carried to the sound gate, where it either leaves the film 
or raises it off the sound gate. 

PRESIDENT CRABTREE: You are speaking of the old method of waxing with 
solid wax? 

MR. GRIFFIN: Yes. 

PRESIDENT CRABTREE: You should use a solution of wax in a solvent. It 
is only necessary to put on a layer of wax a millionth or so of an inch thick, to 
provide the necessary lubrication. 

MR. GRIFFIN: I have seen, in cases where the film is put on a rewinding device, 
two pieces of tallow right at the sprocket holes, over which this film is drawn. 
The projectionist should be taught not to do a thing like that. We must find 
a proper means of treating the filtri so that during projection under high amperages 
it does not seize in the tension parts of the projector. 

PRESIDENT CRABTREE: Of course, Mr. Crabtree is not projecting under the 
high amperages that you speak of. 

MR. FAULKNER: When the film comes off the drying cabinets and is pro- 
jected for inspection, felt runners are used in some places, and I know one labora- 
tory that does not use them. They never scratch film, but it is due to the fact 
that there is no heat on them. 

MR. GRIFFIN: We supply thousands of different types of runners to the 
laboratories of studios, and I know how they work. They use a Mazda lamp, 
and very little light. 

PRESIDENT CRABTREE: I happen to have done a considerable amount of 
research on the lubrication of film. Our researches have shown that if you 
have even the merest trace of wax or oil or grease or any lubricant, on the film, 
it makes a tremendous difference in the ease with which it passes through the 
projector. To date we have not found that any special processing treatment is 
any better with regard to lubrication. 



REPORT OF THE PROJECTION THEORY COMMITTEE 
SUBCOMMITTEE ON LITERATURE 

At the Spring Convention at Hollywood a report was made of 
the activities of the Projection Theory Committee. A subcom- 
mittee to examine the literature of the subject was formed, consist- 
ing of C. Tuttle, F. K. Moss, and H. P. Gage, Chairman. The 
present plan of this Committee is to prepare a tutorial paper on the 
progress of the optics of motion picture projection, based principally 
on the papers published by of the Society of Motion Picture 
Engineers, but also referring to significant papers in other publica- 
tions. 



114 PROJECTION THEORY COMMITTEE [J. S. M. P. E. 

A letter from F. K. Moss states, "I have devoted some time in sur- 
veying the literature on the effect of motion pictures upon the eyes. 
As I progressed in my survey it became increasingly apparent to me 
that available data on the effect of motion pictures upon the eyes 
were largely negative in character. In other words, pictures made 
according to the best of modern practice had little if any observable 
deleterious effect upon the eyes. In the past, when pictures pre- 
sented excessive brightness contrasts, unsteadiness and flicker, -there 
was no doubt that they were the cause of ocular strain and fatigue. 
These objectionable characteristics seem to have been reduced to a 
point where they cease to be important in the better grade of pictures. 

"I also approached the problem more or less directly from the view- 
point of general physiological optics. Such an analysis indicates, 
for example, that the brightness contrasts on the screen and with the 
general surroundings, are not of such an order as to induce unusual 
degrees of ocular fatigue. Hence the conclusions reached by scientific 
considerations and those resulting from actual experience are in agree- 
ment. Since the subject is largely one of eye-strain or 'eye-fatigue' 
which has never been satisfactorily measured, a quantitative discus- 
sion is impossible. In brief, these effects upon the eyes become 
important in cases where projection is faulty." 

Mr. Tuttle under date of April 28 sent a list of forty titles containing 
significant information on the subjects of Illumination, Optics, Pro- 
jection Angle, Mechanics of Projectors, Aberration of Lenses, Pro- 
jection under Special Conditions, and Visual Angle. 

Subcommittee on Literature 
H. P. GAGE, Chairman 
F. K. Moss 
C. TUTTLE 

DISCUSSION 

MR. MATTHEWS: In connection with the work of the Progress Committee, it is 
worthy of mention that a considerable amount of information has been published 
on the subject of visual fatigue in motion picture theaters, in the International 
Review of Educational Cinematography. A series of measurements were made of a 
great many school children in theaters in Italy, giving data that might be con- 
sidered by the Committee. There is a series of four or five articles in this publi- 
cation. 

MR. MURRAY: Does the work of the Committee include a search of the 
literature in regard to the psychological effects involved in the projection of 
motion pictures in color? 

MR. GAGE : Mr. Moss is studying the literature dealing with the effects on the 



Jan., 1932] PROJECTION THEORY COMMITTEE 115 

eye. The Committee is considering more generally the possible deleterious effects 
rather than the whole group of psychological effects, which constitutes such an 
unlimited field that little would be accomplished if it were to be considered in its 
broadest aspect. 

MR. MURRAY: I have heard complaints that colored pictures produce eye fa- 
tigue that I have not heard in connection with black and white. Have these been 
considered by the Committee? 

PRESIDENT CRABTREE : In the case of the old two-color additive Kinemacolor 
pictures, I would say that they caused fatigue. With modern two-color subtrac- 
tive pictures fatigue may have been caused by the fact that some of the pictures 
were out of focus. They lacked definition, and the person viewing them did not 
know whether it was his eye that was at fault or the picture. He assumed that 
his eyes were at fault, and strained them in trying to focus the picture. 

I asked for a vote the other day as to whether the colored pictures we saw one 
evening this week caused any annoyance or eye-strain, and no one said that they 
had any effect they did not seem to notice any difference between the effects 
produced by the colored and by the black-and-white pictures. Perhaps at this 
time, if any one has thought it over and has the courage to say it gave him annoy- 
ance, he might care to say something about it. 

MR. J. CRABTREE: During the showing of the picture I thought I should be 
able to view it to the end, but I had to close my eyes. When the next black-and- 
white picture was projected, the annoyance entirely disappeared. Checking 
with other people, no one else to whom I spoke seemed to have had the same ex- 
perience. Apparently it was merely an idiosyncrasy. But the irritation was 
undoubted hi my case. 

MR. FALGE : Is it not true that with color in general, it is harder to focus on 
some colors than on others, and that one experiences certain visual effects with 
pictures in the blue and red ? 

Another factor which is a function of eye-strain is the size of the picture. The 
magnetoscope pictures, if viewed throughout an entire show, would be very hard 
for those seated in the front rows, as their eyes have to chase back and forth across 
the picture as in a three-ringed circus. Does not the addition of color to the 
picture in general cause a reduction of its brightness? And from the standpoint 
of lighting, it usually f ollows that a decrease hi brightness is less harmful to the 
eyes than an increase. And aren't we back to the same subject we were on a min- 
ute ago, that we have not enough brightness hi our pictures today, and that 
that is harmful to the eyes? 

PRESIDENT CRABTREE: We are talking of annoyance of a much higher order 
of magnitude than what you have in mind. 

MR. BURNETT: In color photography, it has been known for a long time 
that the red colors are very harmful to the eye, while the green colors are not. In 
most of the colored pictures that I have seen the reds have been predominant, caus- 
ing a great deal of strain on eyes which have not been strong enough to stand it. 
I do not think that this effect was as conspicuous the other night as heretofore. 

Eye-strain can be also considered from the standpoint of brightness; but it is 
the reds, I think, that cause the greatest trouble in color photography, as far 
as eye-strain is concerned. 



ABSTRACTS 



Supply and Cost of 16-Mm. Film for the Home. F. S. IRBY. Electronics, 
August, 1931, p. 48. An analysis of the various factors that contribute to the 
cost of 16-mm. films for sound pictures in the home. The author considers that 
if such films are to reach more than a very limited class market, the rental cost to 
the consumer should not exceed $2 for a four or five reel feature picture. The 
library must anticipate liquidation of the cost of the film in from twenty to 
twenty-five rentals, which means that the cost of the film to the library must not 
exceed $10 to $12 per reel. A. C. H. 

Light-Valve Sound Recording. J. P. LIVADARY. Electronics, August, 1931, 
p. 54. The third and final installment in a series of articles dealing with the fre- 
quency distortion introduced by the finite width of the slit in recording. This 
article is concerned chiefly with a mathematical analysis of the distortion intro- 
duced in the light valve method of recording. In the conclusion, the author 
summarizes the results of this and the preceding articles by comparing the various 
methods of recording that have been studied; namely, the glow lamp method, 
the single ribbon light valve method, the double ribbon light valve, and the 
variable width method. He concludes that "from a practical standpoint, all 
three systems are capable of high-grade recording, and any difference such as we 
have shown will not become very apparent or objectionable until such time when 
the film grain noise is suppressed and sound recording systems are capable of 
commercially reproducing frequencies up to 10,000 cycles or over. Until then all 
three systems will be competing on practically equal terms." A. C. H. 

Dynamic Loud Speaker Design. J. E. GOETH. Electronics, August, 1931, 
p. 66. A very elementary account of the magnetic circuit of dynamic loud speakers. 
A second installment of this article will appear in a later issue. A. C. H. 

A Rapid-Record Oscillograph. A. M. CURTIS AND I. E. COLE. Electronics, 
August, 1931, p. 70. An oscillograph of the string galvanometer type that is 
especially designed for the study of transient phenomena. A. C. H. 

Noiseless Sound-on-Film Recording. GEORGE LEWIN. Electronics, Septem- 
ber, 1931, p. 102. The author discusses the theory of noiseless sound-on-film 
recording by the light valve. The subject will be treated from a practical stand- 
point in a subsequent issue. A. C. H. 

Dynamic Loud Speaker Design II. J. E. GOETH. Electronics, September, 
1931, p. 112. The second and final installment of an article concerned primarily 
with the magnetic circuit of dynamic loud speakers. A. C. H. 

Stage Equipment: An Outline of Modern Practice. W. L. TANN. Theater 
Management, 27, December, 1931, p. 6. Essential stage equipment in an average 
sized theater presenting straight pictures or pictures and stage performances is 
described and illustrated. Modern advances in fire protection by asbestos cur- 
tains and steel smoke pockets are pointed out. Various mechanical contrivances 
for enlarging the screen to permit the showing of wide films are discussed. A 
116 



ABSTRACTS 117 

notable advance in design of stage equipment is the silence with which the intri- 
cate mechanism operates. E. P. J. 

Room Noise Reduction for Improved Sound Reception. V. A. SCHLBNKBR. 
Theater Management, 26, November, 1931, p. 3. Describes tests conducted to 
determine the effect of extraneous noises on sound reproduction. Illustrates out- 
side noises in typical theater before and after acoustical treatment of vestibule, 
lobby, foyer, and exit doors. An oscillograph trace of three bands of noises re- 
corded simultaneously in the street, lobby, and foyer of theater under discussion 
reveals that while high and middle frequency bands are effectively silenced by 
entrance doors, bands of low frequency enter the theater practically undiminished. 

A chart showing the effect of various sensation levels expressed hi decibels 
above minimum audibility of the human ear is discussed. The painful effect 
produced by fader manipulation to produce audibility of picture sound above 
room noise is indicated. E. P. J. 

The Use of Rochelle Salt Crystals for Electrical Reproducers and Micro- 
phones. C. B. SAWYER. Proc. IRE, 19, No. 11, November, 1931, p. 2020. A 
brief history of the use of piezo-activity for acoustic uses is followed by a descrip- 
tion of a cheap method of production of Rochelle salt crystals, used in the author's 
experiments. The principle of opposition was used. Two Rochelle salt sec- 
tions are cemented together so that upon application of an electrical field, one 
section tends to expand and the other section tends to contract, thus amplifying 
the resultant motion. The method of cutting Rochelle salt crystals for this work 
is explained. Brief descriptions of Rochelle salt microphones, loud speakers, and 
phonograph pick-ups are given. The Rochelle salt development has the follow- 
ing outstanding advantages. 

(1) Cheapness and simplicity. 

(2) Long life. 

(3) Flexibility of design. 

(4) Generation of high voltages in input circuits. 

(5) Directly matched with output tubes in output circuits. 

(6) No necessity for an exciting field. A. H. H. 

Trans-Lux Rear Stage Projection. W. MAYER. Theater Management and 
Theater Engineering, 26, No. 22, October, 1931, p. 3. A non-technical discussion 
of the Trans-Lux system of rear stage projection as installed in theaters. The 
history of the system, various problems encountered and their solutions, and a de- 
scription of the present installations give a concise outline of Trans-Lux. By 
means of special lens and optical systems, no changes in the projector and sound 
head mechanisms are necessary. Standard film is used and is threaded in the 
projector in the standard way. The average distance between screen and pro- 
jector is 13V2 feet. A - H - H - 

Moving Coil Telephone Receivers and Microphones. E. C. WENTE AND A. L. 
THURAS. Bell Telephone Tech. J., X, No. 4, October, 1931, p. 565. A descrip- 
tion of a moving coil head receiver and a microphone. The mechanical construc- 
tion is based on using light-weight materials for moving parts, thus giving greater 
response over the frequency range. Theoretical and actual response are com- 
pared. The sensitivity of the moving coil microphone was found to be about ten 
db. higher than that of the condenser microphone. A. H. H. 



1 18 ABSTRACTS [J. s. M. P. E. 

Playing Light on a Thermionic Organ. W. C. FULTON Motion Picture 
Herald, 104, No. 13, September 26, 1931, Section 2, p. 12. A description of a 
unique lighting switchboard, built for the Severance Memorial Hall in Cleveland. 
The major innovation in the lighting system is the switchboard, built along the 
lines of a console of a modern organ. Controls for 4000 lighting combinations of 
110 load circuits are at the finger tips of the operator. Included are a four scene 
preset control, proportional control, remote control of intensity, and inter- 
connection of circuits. The system is based on the thermionic type of lighting 
control. The control apparatus for each circuit requires a dimming reactor, a 
conventional vacuum tube, two grid glow rectifiers, and a system of control po- 
tentiometers. The lamp load current flowing in the a-c. coils of the reactor, is 
directly dependent on the d-c. saturation current flowing in the d-c. coil of the 
same unit. As the direct current increases, the iron core of the reactor becomes 
saturated alternating current increases. The direct current is supplied by a pair 
of grid glow tubes whose output is controlled by the plate current of the vacuum 
tube. The plate current of the vacuum tube is in turn controlled by varying 
the bias on its grid. All the above apparatus is placed at a remote point from 
the control console. The control circuit of the vacuum tube grid is brought to the 
console. By means of selector switches, potentiometers, etc., any or all circuits 
in the hall may be controlled at will. Circuit diagrams and pictures clearly show 
the operation of this installation. A. H. H. 

Audible Frequency Ranges of Music, Speech, and Noise. W. B. SNOW. Bell 
Telephone Tech. J., X, October, 1931, No. 4, p. 616. A description of tests to 
determine the maximum frequency range necessary for perfect or nearly perfect 
reproduction. With the aid of experienced listeners, and using a series of filters, 
varying degrees of cut-off were tried. It was found that frequencies between 80 
and 8000 cycles were necessary to give good quality. Although rather indefinite 
as to the advantages of using frequencies outside this range, it is believed that the 
most nearly perfect quality is obtained by reproducing the full audible frequency 
range. A. H. H. 

The Development of the Microphone. H. A. FREDERICK. Bell Telephone 
Quarterly, July, 1931, p. 164. An interesting history of the early experiments 
leading up to the present design of microphones. Dr. Page in 1837, Sullivan in 
1845, Bourseil in 1854, Reis in 1861, Helmholtz in 1863, and Varley in 1870, made 
contributions to the development of the microphone. The experiments of 
Dr. Alexander Graham Bell, begun in 1874, are described in more detail. In 
1877, Edison patented a transmitter of the varying resistance type, using a button 
of solid carbon or plumbago. The granular carbon design was first used in 1885. 
The condenser type and the piezoelectric crystal type are of more recent design. 
The difficulties of developing the carbon microphone are described in detail. It 
is interesting to note that minute granules of carbonized anthracite coal were first 
used by Edison in 1886. This source of carbon is still used to a great extent at 
the present time. A. H. H. 

The Effect of Humidity upon the Absorption of Sound hi a Room, and a De- 
termination of the Coefficients of Absorption of Sound in Air. V. O. KNUDSEN, 
JR. /. Acoustical Soc. of America, HI, No. 1, Part 1, July, 1931, p. 126. It is 
shown that the absorption of sound hi air for frequencies above 2000 cycles is ap- 
preciable. This effect is great enough to affect very appreciably the calculation 






Jan., 1932] ABSTRACTS 

of the reverberation time and absorption in a room for frequencies of 4000 cycles 
and above. The absorption of air becomes less as the humidity increases. 

An idea of the magnitude of the effect may be obtained from the following 
statement. "Thus, if a tone of 4096 d.v., in the form of a plane parallel beam, 
were used for long range signaling there would be, at a temperature of 21 C. and 
a relative humidity of 44 per cent, an attenuation of 9.8 db. per second, or about 
46 db. per mile. On the other hand, the attenuation would be less than 1 db. per 
mile for a frequency of 512 d.v." Furthermore, it appears from this data, that a 
reverberation chamber with perfectly reflecting walls would have a reverberation 
time of no more than about six seconds for a tone of 4096 d.v. if the humidity of the 
air in it is 44 per cent or less. 

Theoretical formulas are deduced. The method used in separating the effect of 
the absorption in air and that at the surface of the rooms was to take comparable 
data in two rooms of different sizes but with the same boundary material, namely, 
painted and varnished concrete. This yields sufficient data to separate the effects. 
Even at 4096 d.v. the absorption of the painted concrete was about 0.02 and prac- 
tically independent of humidity as long as condensation did not occur. 

W. A. M. 

A Critical Study of the Precision of Measurement of Absorption Coefficients by 
Reverberation Methods. P. E. SABINE, JR. /. Acoustical Soc. of America, HI, 
No. 1, Part 1, July 1931, p. 139. The data presented include a comparison of 
absorption coefficients obtained at the Bureau of Standards and by two methods 
at the Riverbank Laboratories on identical samples of each of four materials. 
It is concluded that normal experimental errors in measuring absorption coef- 
ficients may easily be 3 or 4 per cent, that probably an error of 10 per cent in the 
coefficients would not appreciably affect the acoustic properties of an audience 
room; and the actual computation of the reverberation time in a room is a matter 
of approximate estimate rather than precise determination. W. A. M. 

The High Intensity Arc for Motion Picture Projection. F. PATZELT. Kino- 
technik, 13, September 20, 1931. p. 344. Measurements and graphs were made of 
the light distribution of an "Artisol 75" projection lamp with high intensity car- 
bons and with ordinary carbons. The average brightness of the entire crater of 
ordinary carbons 14 mm. in diameter at 35 amperes and 45 volts was found to be 
140 Hefner candles per sq. mm. Copper-coated high intensity carbons 11 mm. 
in diameter were found to have a brightness of 357 Hefner candles per sq. mm. 
at 75 amperes and 45 volts. The variation in the brightness of high intensity 
carbons with different amounts of current was also measured. It was found that 
carbons of small diameter require higher current densities than larger carbons to 
attain the same brightness. The effect of changing the relative positions of the 
carbons was studied, and it was found that greater brightness was attained with 
the axis of the negative carbon in line with the center of the positive carbon than 
with the axis of the negative carbon opposite the lower edge of the positive carbon. 
The variation of the brightness at constant current with varying length of arc was 
found to be small. It is stated that a 25-degree inclination of the axis of the nega- 
tive carbon to the axis of the horizontal positive carbon is the most favorable. It 
is concluded that the difficulties in the use of high intensity carbons are compen- 
sated for by the increased illumination. M. W. S. 

Safety Film. K. BRATRING. Kinotechnik, 13, July 20, 1931, p. 237. In its 



120 ABSTRACTS []. S. M. P. E. 

mechanical properties, such as resistance to wear and damage, cellulose acetate 
motion picture film base is considered inferior to cellulose nitrate base. In view 
of the universal precautions against fire in the projection of professional motion 
picture films, it is considered that the low inflammability of cellulose acetate film 
is sufficient cause to justify the increased expense attendant upon its use in thea- 
ters. For schools, homes, and other places where proper safety precautions for 
nitrate film are not taken, cellulose acetate film should undoubtedly be used. It 
is thought that nitrate support constitutes no great hazard when used for amateur 
roll films and film packs, or for professional portrait films. For x-ray films, the 
introduction of cellulose acetate support is viewed with favor. M. W. S. 

The Phillips Reproducing Set. Kinemat. Weekly, 172, June 4, 1931, p. 61. 
The sound equipment in the Phillips set is a pedestal mounted at the left-hand side 
of the projector; and a flexible shaft coupling driven by the motor is connected with 
the projector flywheel. An integral gear shift permits the use of either sound-on- 
film, sound-on-disk, or silent operation. The sound head of the projector em- 
ploys a curved gate which is said to prevent film buckle. A high emission photo- 
electric cell (18 microamperes per lumen) is used at present but a gas-filled 
caesium cell is being investigated for future use. The speed control is ingenious, 
the electric control being effected by rotating make-and-break cams, one driven 
by the projector motor and the other by a constant-speed motor. When the 
contact is made on both cam switches, a resistance is short circuited. The 
period during which this resistance is short circuited, therefore, depends upon the 
relative positions of the two cams. The cams revolve at approximately 80 rpm. 
The fader used in the set gives a logarithmic change. The projection room ampli- 
fier consists of a single stage which supplies current to the main amplifier which 
may range in capacity from 20 to 200 watts with speech levels of 10 to 45 watts, 
respectively. L. E. M. 

A Continuous Motion Picture Projector. M. Hue. Bull. soc. frang. phot., 
73, June 1931, p. 128. A newly designed single oscillating mirror type of con- 
tinuous projector is described. The principle involved is one in which the film 
passes over a cylindrical drum having an aperture through which the single frame 
is projected upon an oscillating mirror, which in turn reflects it into the objective 
of the machine. During the movement of the film over the aperture, the adjacent 
frame is isolated by a moving window behind the aperture, which moves with the 
same angular velocity as the film. When the projection phase is terminated, a 
shutter in front of the objective masks it during the return of the mirror and win- 
dow. The light from the illuminating sources does not fall directly on the film 
but is interrupted and reflected by a 45-inch mirror which is fabricated of a metal 
capable of absorbing a large percentage of the heat rays, thus protecting the film. 
All gears and cams are encased in oil, where possible, thereby minimizing noise. 
It is claimed that a projector as described is capable of projecting a film 3000 
times without injury to the film. Drawings are included. C. H. S. 

Faith in the Title. F. SLIP. Filmtechnik, 7, May 2, 1931, p. 6. Although 
titles have been replaced temporarily by the use of sound, they have a place in 
certain classes of films, such as teaching films. Correctly composed titles may 
also be of value in the presentation of certain sound films. During a study of 
correct methods of title composition the maximum title width of 19 mm. has been 
selected as desirable with the height accordingly proportional. The background 



Jan., 1932] ABSTRACTS 121 

should preferably be dark and the letters light. The type must be simple, clear, 
and attractive. The optimum length for the title has been investigated from a 
consideration of (1) length of the lines, and (2) number of letters. A useful table 
is given showing length of the lines, number of letters, length of the title, and 
length of the film per line of title, assuming projection at the rate of 24 frames per 
second. L. E. M. 

Motion Picture of the Eclipse of the Moon. F. Albrecht. Filmtechnik, 7, May 
2, 1931, p. 1. On April 2, 1931, the first motion picture of a total eclipse of the 
moon was photographed at the Trepton observatory hi Germany. With the 
usual motion picture camera the image of the moon is far too small and even with 
a teleobjective of 30 cm. The image is only 3 mm. in diameter. In the successful 
motion picture an //10 objective of 65 cm. focal length was used, mounted on an 
Ernemann E camera. The camera and lens were secured in place on the 21-meter 
Trepton telescope. Positive film was employed and exposures of l / 4 to Vz second 
were made, using a blue filter with the teleobjective operated with a 35-mm. open- 
ing. The camera shutter opening was increased to 160 degrees at the beginning of 
the eclipse and decreased to 90 degrees as the eclipse passed. Single frame ex- 
posures were made at intervals of 5 seconds, thus giving for the 3 ! A hour time a 
length of film which, when projected at the rate of 24 frames per second, occupied 
P/2 minutes. L. E. M. 

Television Demonstration at Broadway Theater. Film Daily, 57, October 23, 
1931, p. 1. A television demonstration was given at the Broadway Theater, 
New York, for two weeks beginning on Oct. 22, 1931, a 10 by 10 foot screen being 
used. The receiving disk revolved 900 times per minute and a projection system 
projected the images on the screen. The sending station was located a short dis- 
tance away in the Theater Guild Studio. G. E. M. 



BOARD OF ABSTRACTORS 

CARRIGAN, J. B. MACFARLANE, J. W. 

COOK, A. A. MACNAIR, W. A. 

CRABTREE, J. I. MATTHEWS, G. E. 

FOWELL, F. McNicoL, D. 

HAAKE, A. H. MEULENDYKE, C. E. 

HARDY, A. C. MUEHLER, L. E. 

HERRIOT, W. PARKER, H. 

IRBY, F. S. SANDVICK, O. 

IVES, C. E. SCHWINGEL, C. H. 

KURLANDER, J. H. SEYMOUR, M. W. 

LOVELAND, R. P. WEYERTS, W. 



ABSTRACTS OF RECENT U.S. PATENTS 

The views of the readers of the JOURNAL relative to the usefulness to them of the 
patent abstracts regularly published in the JOURNAL will be appreciated. Favorable 
views are of particular interest. In the absence of a substantial body of opinion 
to the effect that these patent abstracts are desired by the membership, their early 
discontinuance may be considered. 

1,821,930. Film Feeding Mechanism. M. COUADE. Sept. 8, 1931. A film 
feeding mechanism for projectors in which a claw engages the perforations in the 
film and intermittently moves the film in accordance with the operation of a 
cam mechanism which imparts angular movement to the claw. Adjustments 
may be made for determining the length of stroke of the claw by adjusting the 
eccentricity of the driving cam mechanism which engages the claw. 

1,821,946. Film Feeding Mechanism. F. H. OWENS. Sept. 8, 1931. A 
sound and motion picture apparatus including mechanism for intermittently 
moving the picture films in front of the projection lens system while continually 
moving the sound record portion. The shutter for the light beam has a peri- 
pheral groove thereon for defining a belt wheel which is engaged by the drive belt. 
A manual adjusting means is provided for properly positioning the shutter. There 
is a lost motion connection between the film moving mechanism and the parts of 
shutter by which the shutter may be selectively adjusted under manual control 
before being operated under automatic control. 

1,822,057. Composite Photographic Sound Records. F. H. OWENS. Sept. 
8, 1931. The method of making a composite photograph sound record from dif- 
ferent sources of sound such as a song with orchestra accompaniment and with 
the addition of some special instrumental features such as bells and the like where- 
in a plurality of photographic sound records are synchronously converted into 
electric impulses. These impulses are received for modulating the intensity of a 
single recording lamp. The modulated light rays from the lamp are photographed 
upon the sensitized film. By this process it is possible to produce a sound record 
by selecting desirable portions of previous sound records and thereby construct a 
program of highly entertaining qualities. 

1,822,183. Light Slit for Recording and Reproducing. D. A. Whitson. As- 
signed to Whitson Photophone Corp. Sept. 8, 1931. A light slot for a sound 
recording and reproducing system in which a guide block is disposed adjacent the 
film. The guide block has a wide slot and a communicating narrow slot. The 
sound record is passed over the wide slot. There is a lens in the bottom of the 
wide slot and in contact with the sides and bottom of the slot for focusing radia- 
tions to pass through the slots on the strip. The purpose of the lens slot is to con- 
centrate the light at maximum intensity upon the film at the same tune that pro- 
tection of the slot against the accumulation of dust or foreign matter is effected. 

1,822,350. Arrangement of Perforations in Cinematographic Films. J. H. 
JARNIER. Sept. 8, 1931. A motion picture film which is perforated laterally of 
the picture frames instead of in two rows on opposite sides of the picture frames. 
122 



PATENT ABSTRACTS 123 

Claws are used to shift the picture frames intermittently before the projector. 
The structure of the film is such as to increase the resistance of the film against 
tearing at the lateral lines of perforations. Rectangular perforations are provided 
in the transverse spaces between the images wherein the ratio of the number ( JV) 
of transverse perforations to the width (Z,) of the film having a specific resistance 
to rupture by traction X is determined by the formula: 



p a\ 

in which a is the width of the perforations, p the resistance to rupture for the width 
a in such a way that the resistance to tearing of the line of perforations engaged 
is the same as the resistance to rupture by traction of the spaces separating them, 
this resistance being the maximum. 

1,822,528. Moving Lens Cinematograph Machine. W. E. JOHN. Sept. 8, 
1931. A motion picture camera or projector having a continuously moving film 
and a series of loose lens carriers moving with the film. The loose lens carriers 
move through a closed circuit including a straight guide in which they are exposed, 
and curved guides, one at each end of the straight guide; the circuit between the 
curved guides being completed by a driving and conveying member in the form of 
an internally toothed and pocketed wheel. The separate lens members are 
brought into alignment with the optical path through the camera by driving 
means connected with the lens carrier. The lens carriers slide longitudinally 
around the guide which defines the path of movement for each of the lens members. 

1,822,551. Lens Shifting Mechanism for Projecting Machines. A. TON- 
DREAU. Assigned to Warner Bros. Pictures, Inc. Sept. 8, 1931. A system of 
lenses which may be shifted in a projection machine to enable an instant change of 
magnification on the projection screen without loss of focus. An attachment is 
provided carrying lenses which may be first set in focus and which may be operated 
to bring either one lens of a certain magnification or another lens of a different 
magnification into the optical path. The lens carrier is provided with individual 
supports for the different lens members, allowing independent longitudinal ad- 
justment of the different lens carriers. 

1,822,865. Glow Discharge Tube for Recording. T. W. CASE. Sept. 8, 1931. 
A glow discharge tube for recording variations in light intensity upon film. A 
bulb is provided for enclosing a non- thermionic anode and a cathode. An atmos- 
phere of helium is provided within the bulb at such a pressure that a concentrated 
glow is provided about the negative electrode with a voltage not substantially 
greater than 400 volts d-c. across the electrodes. The cathode has a photoelec- 
trically activated coating comprising barium actuated for electron emission by the 
said glow concentrated about the cathode. The device is designed to produce 
response of the glow in terms of light emission according to variations of electrical 
impulses produced in a sound control circuit. 

1,822,932. Combination Recording and Reproducing Stylus Head. M. H. 
LOUGHRIDGE. Sept. 15, 1931. A stylus head is arranged to support both a 
recording and a reproducing stylus with respect to a wax record of a phonograph. 
The stylus head may be shifted to bring either the recording or reproducing stylus 
into engagement with the phonograph record. A switching mechanism is pro- 
vided for controlling the connection of the styluses to an amplifying system. 



124 PATENT ABSTRACTS [J. S. M. P. E. 

When the reproducing stylus engages the sound record, the input circuit of the 
amplifier is connected with the stylus. When the recording stylus is placed in 
engagement with the sound record, the magnetic windings thereof are connected 
with the output circuit of the amplifier for cutting a groove in the record in ac- 
cordance with the sound vibrations impressed upon the input circuit of the ampli- 
fier. 

1,823,243. Method and Apparatus for Lapping Color Film Embossing Rollers. 
O. WHITTEL. Assigned to Eastman Kodak Co. Sept. 15, 1931. A method of 
lapping lenticular film embossing rollers which comprises providing a cylinder with 
a plurality of fine guide lines, turning the cylinder, and lapping the cylinder 
with a plurality of wires, a fine lapping compound being used on the cylinder. The 
embossing roller is used for operation upon color motion picture films. The len- 
ticular areas or elements formed in the film are extremely minute as the distance 
across these elements may be from 0.0015 to 0.002 of an inch. 

1.823.245. Film Winding Device. O. WITTEL. Assigned to Eastman Kodak 
Co. Sept. 15, 1931. Winding device for motion picture film in which a reel 
is provided with a pair of concentric hub members. One hub member is slidably 
carried by a flange disposed in one side thereof. The two hub members are sepa- 
rated by sliding the flange on one hub. The structure of the film winding device 
is such that the film may be drawn from an inner convolution of a supply reel and 
wound on an outer convolution of a take-up reel. The construction of the reel is 
such that the film is properly aligned on the reel without rewinding. 

1.823.246. Method of Tinting Film for Use in Sound Reproduction. A. A. 
YOUNG. Sept. 15, 1931. A method of tinting the picture areas of a photograph 
film in which the sound record portion is preserved untinted while preventing 
shrinkage of the film by applying to the picture areas of the film a dye dissolved 
in a solution comprising a solvent for the film and the dye and a non-solvent for 
the film which has the property of reducing the rate of evaporation of the solvent 
whereby the tendency of the film to buckle is eliminated. The dye, which is 
applied to the picture areas of the film, is dissolved in a solution containing from 5 
to 10 per cent of acetone, from 70 to 75 per cent methyl alcohol, and the remainder 
triacetin. 

1,823,349. Producing Fade-in and Fade-out of Photographic Sound Record. 
S. C. CHAPMAN. Assigned to Electrical Research Products, Inc. Sept. 15, 1931. 
The sound record is chemically treated for reducing the end portions of the sound 
record progressively varying lengthwise of the film. The reproduced sound will 
thus gradually increase in volume from silence to the normal volume of the 
record, vary normally with the record till near the end when the volume of 
the sound will gradually diminish to silence. 

1,823,355. Telescope Framing Device. L. S. FRAPPIER AND E. BOECKING. 
Assigned to International Projector Corp. Sept. 15, 1931. Projecting machine 
for photographic sound records wherein a microscope is supported in the path of 
a scanning ray in such position that the ray can be observed while adjustments 
are being made to secure the proper characteristics thereof. A prism is positioned 
in the path of the light rays to deflect a portion of the light at right angles into 
the microscope in order that the sound record may be analyzed. 

1,823,400. Photographic Film Copying Machine. L. HORST. Assigned to 
Sinus Kleuren-Film Maatschappil, of Bosch en Duin, Netherlands. Sept. 15, 



Jan., 1932] PATENT ABSTRACTS 125 

1931. A machine for copying two color films and more particularly a machine 
of this kind in which the pictures are transferred from one film to the other by 
means of mirrors and objectives provided in duplicate. Two sources of light are 
provided, each of which is separately regulable, for timing the degree of copying of 
the individual part pictures. 

1,823,462. Photographic Camera. K. MORSBACH. Assigned to Siemens & 
Halske, Aktiengesellschaft. Sept. 15, 1931. The film refill which is supplied 
for the camera is carried by an interchangeable cassette which cooperatively en- 
gages a film guide channel located in the interior of the camera behind the objec- 
tive lens. There is a guide plate carrying the window for the image, permanently 
located behind the objective and in its focus. There is a pressure plate indepen- 
dently mounted on the cassette. When the camera is refilled, any differences in 
the focal lengths of the objectives of different cameras are compensated by the 
pressure plate and the guide plate so that equal operation of cameras which are 
not uniform is obtainable. 

1,823,737. Sound-on-Disk Motion Picture Projector. CHARLES L. HEISLER. 
Assigned to General Electric Co. Sept. 15, 1931. A motion picture projector 
which includes a projector housing mounted adjacent a phonograph turntable. 
The driving motor which operates the projector also drives the phonograph turn- 
table so that the film and the record may be operated in synchronism. The arm 
which carries the phonograph pick-up is pivoted adjacent one side of the record 
table and permits the phonograph pick-up to be moved over the area of the 
revolving record. 

1,824,294. Sound and Picture Film Matching Means. FREEMAN H. OWENS. 
Assigned to Owens Development Corp. Sept. 22, 1931. A method which permits 
the accurate repair or splicing of separate film strips, one of which carries the pic- 
ture record and the other of which carries the sound record to maintain synchro- 
nism between the picture and the sound wherein an insertable film section is pro- 
vided attachable to the broken ends of the film. The insert is provided with a 
sound record and images adjacent the sound record, the images being partial 
duplicates of the images on the picture film. The splicer finds it very easy and 
convenient to judge accurately the length of the insert by simply matching the 
two films by observing the partial duplicates of the images on the insert and fitting 
the sound strip in to match the sound on the film. That is to say, a guide is pro- 
vided on the insertable sound strip so that the splicer is advised accurately as to 
where this sound should occur on the sound film in order to match accurately the 
images on the picture film. 

1,824,417. Treating Sound Records Produced by Splicing. A. T. TAYLOR. 
Assigned to Metro-Goldwyn-Mayer Corp. Sept. 22, 1931. The method of 
splicing a film carrying a sound record to prevent audible clicks and foreign noises 
at the splice marks as the film passes the light path. The ends of the broken film 
are cemented and then a patch in the form of a half -cycle sine wave cemented over 
the adjoining ends of the sound record. This sine wave patch has a frequency 
below normal audibility and an amplitude equivalent to the width of the sound 
record so that there is no extraneous sound created as the splice passes the sound 
reproducing aperture. 

1,824,446. Producing Motion Pictures in Color. E. L. PEARSON. Sept. 22, 
1931. A projection screen is arranged for rotative movement in timed relation tQ 



126 PATENT ABSTRACTS [J. S. M. P. E. 

the rotation of a color filter at the projection machine. The projector is arranged 
to project successively images through the different colored filters upon the moving 
projection screen from which the picture may be viewed and through which the 
images are produced. By shifting the relative positions of the projector and the 
projection screen to project successively the images upon the portions of the pro- 
jection screen corresponding to the particular filters upon which the images are 
produced, an effect upon the eye of colored motion pictures closely portraying in 
color and motion real animated objects is produced. 

1,824,709. Camera for Taking Cinematographic Pictures. A. L. V. C. DE- 
BRIE. Sept. 22, 1931. View taking apparatus comprising two parts, namely, a 
front part containing the film driving device, the shutter, and the optical arrange- 
ment and a rear removable part which can be secured instantaneously to the front 
part and which contains a feeding storing box wherein the unimpressed film is 
disposed together with the film guiding devices, the transmission gear, and a second 
storing box into which the impressed film is wound up. The latter box can be 
the same as the feeding box or else both boxes can be made separate. The opera- 
tor can thus be provided with several rear parts ready for use which he may se- 
cure to the front part of the apparatus according to the requirements. The result 
thereof is, besides the advantage already mentioned, a saving of time which is of 
great interest in the case where the taking of the complete scene which is to be 
cinematographed requires a length of film greater than what can be contained in 
one single storing box. 

1,824,731. Picture Transmitting System. D. M. MOORE. Assigned to 
General Electric Co. Sept. 22, 1931. A picture receiving system in which the 
light is modulated in accordance with the shading of the successive elemental 
areas of the picture transmitted. A screen is provided and there are a plurality 
of rotatably mounted mirrors arranged to reflect successively the modulated light 
to produce a trace on the screen. The mirrors are rotated continuously in one 
direction at different speeds with a lens system arranged between the mirrors. 
The mirrors are each mounted on the shaft of the associated driving means in 
such manner that the mirrors are inclined at an angle to the axis of the driving 
shaft so that rotation of each of the mirrors produces a scanning operation over 
the area of the receiving screen. 

1,825,078. Incandescent Electric Lamp for Projection Apparatus. J. MA- 
RETTE. Assigned to Pathe Cinema Anciens Etablissements Pathe Freres. 
Sept. 29, 1931. A glow lamp is directly centered in the optical path of a projection 
machine by means of a ring member which is secured over the base of the lamp and 
serves to center the lamp accurately in its support for accurately directing the 
maximum amount of light through the projection path. 

1.825.121. Lamp Holder. F. H. OWENS.. Assigned to Owens Development 
Corp. Sept. 29, 1931. A plurality of separate lamps are mounted on a carrier 
which may be laterally shifted to move any one of the lamps successively into a 
predetermined operative position. There are stops provided on the lamp sup- 
porting base to limit the movement of the lamps to selected positions. The lamp 
holder may be moved through a shaft member to the outside of a lamp housing. 

1.825.122. Objective for Color Photography. A. OSWALD. Assigned to Kel- 
ler Dorian Colorfilm Corp. Sept. 29, 1931. An objective lens system for color 
photography employing films having a goffered base wherein the lens system is 



Jan., 1932] PATENT ABSTRACTS 127 

made up of a plurality of different elements; a diaphragm and a collimator film. 
The several elements of the optical system are so arranged that the pupil of emer- 
gence of the objective is in the anterior focal plane of the collimator lens, the 
aberrations introduced by the collimator lens being corrected by compensating 
aberrations introduced into said objective. 

The objective is anastigmatic and is constituted by the three spaced elements 
and by a collimating lens located in the vicinity of the focal plane of the objective. 
The objective of this invention follows Petzval's law 



and in calculating these objectives in view of increasing the sharpness of the mar- 
ginal images, P is left with a negative value suited to the extent of the field to be 
represented; when calculating an objective of this sort intended to be provided 
with a collimating lens, the residual value ascribed to P will therefore have to be 
increased by varying the quantity <f>rj. 

1,825,142. Motion Picture Film Magazine. W. A. BRUNO. Assigned to 
Clarence W. Fuller. Sept. 29, 1931. A protective housing for films wherein the 
film is supported for avoiding breakage or other injuries. A plurality of film 
carrying reels of considerable diameter are provided so that the film may be 
stored in the magazine, without sharp bends. The reels are constructed to engage 
the film near its marginal edges only, the cylindrical surfaces of the reels being 
concave or otherwise centrally disposed to prevent contact thereof with the 
central portions of the film. Power means are provided for driving the reels for 
storing the film in the magazine while preventing scratching or other abrasion to 
the picture frames on the film. 

1.825.253. Synchronous Camera Mechanism. A. F. VICTOR. Sept. 29, 1931. 
A camera having means for controlling and synchronizing the motion and arresting 
the movement of the feeding devices with respect to the shutter. A cam co- 
operating with an abutting arm is provided in association with the rotatable 
shutter by which the shutter may be brought to rest by moving the arm. By 
withdrawing the arm from the path of the cam the shutter may be rotated under 
control of the drive mechanism. The movement of the shutter with the film 
feeding devices is synchronized. The shutter is provided with additional devices 
that cooperate with control mechanism so that when the latter is released to return 
to normal, the stoppage of the film is momentarily postponed until the shutter 
is in position in front of the exposure aperture, whereupon the movement of all 
mechanism is arrested. This is accomplished in such manner that it positively 
insures the proper positioning of the shutter in front of the aperture at the moment 
the movement of the film ceases and the stoppage is made without jar to the 
camera. 

1.825.254. Intermittent Feed for Motion Picture Apparatus. A. F. VICTOR. 
Sept. 29, 1931. A mechanism for intermittently feeding a film through a camera 
or projection machine which includes a shuttle that is reciprocated by a continu- 
ously rotatable cam. The shuttle is hinged upon the end portions of lever arms 
that are pivotally mounted upon the housing of the camera or projector. Means 
are provided for adjusting the pivoted ends of the arms toward each other in 
such manner that any noticeable wear between the cam and the parts engaged 



128 PATENT ABSTRACTS [J. S. M. P. E. 

thereby may be taken up by means of a simple adjusting structure. The ful- 
crums of the lever arms are supported in a "floating" pivot because the pivotal 
members are not actually secured to the camera or projector but are carried upon 
suitable rocker-arms which themselves are pivoted on the support or housing. 
The operation of these rocker arms is similar to the action of a cam or cams en- 
gaged with the lever arms. 

1,825,340. Electrooptical Cell. N. DEISCH. Sept. 29, 1931. A Kerr cell is 
used for electrically modulating a beam of light. One electrode of the Kerr cell 
comprises a frame having an opening comprising the active space thereof and a 
plurality of flexible ribbon-like division members dividing the opening into a 
plurality of light passages, the flexible ribbon-like members being secured to said 
frame and held taut across said opening. Electrostatic stresses impressed on the 
cell operate to modify the light passing through the divisions of the cell. 

1,825,529. Sound Pipe Reproduction from Photographic Films. R. KOLLER. 
Sept. 29, 1931. A motion picture film is combined with an air control band which 
moves synchronously with the motion picture film. The ah" control band moves 
over a tracker board for controlling the supply of air to various sound pipes for 
the reproduction of sound appropriate to the pictures. Synchronization of the 
sound with the pictures is obtained by virtue of the interconnection of the moving 
band with the picture film. Various forms of pipe organ valves may be oper- 
ated by allowing the air to pass through predetermined apertures hi the moving 
band. 

1,825,486. Scanning Disk. A. O. TATE. Sept. 29, 1931. The apertures in 
a scanning disk are arranged in reverse spirals, one of the spirals beginning at the 
outer edge of an image and ending at the inner edge thereof and the other of the 
spirals beginning at the inner edge and ending at the outer edge. The adjacent 
apertures of the spirals are disposed the same radial distance from the center 
of the disk so that the images are scanned twice in succession. Each of the aper- 
tures is bounded by arcs of concentric circles and by radii of the disk. The ob- 
jects of the arrangement of the scanning disk apertures are to eliminate the incon- 
venient restrictions with respect to the area available for use as scanning space as 
defined by the distances between the open ends of the spirals, to provide means 
whereby an object may be scanned laterally by intermittent light beams or pen- 
cils which maintain perpendicularly a continuous, rhythmic, undulatory movement 
through the period of revolution of the disk; to provide means whereby the total 
area of the scanning space may be varied with respect to its dimensions; and to 
provide means whereby an object may be scanned with one revolution of the disk 
a plurality of times. 

The scanning disk is divided circumferentially by a plurality of radial lines to 
form circumferential divisions and is divided radially by a plurality of concentric 
circles to form radial divisions and may be conveniently plotted by the following 
formula, in which: 

A represents the number of circumferential divisions of the entire disk; 

B represents the number of radial divisions included within the scanning area; 

C represents the number of circumferential divisions between successive aper- 
tures; and 

D represents the number of times the scanning area is scanned hi one revolution 
of the disk and also the number of spirals in the system. 



Jan., 1932] PATENT ABSTRACTS 129 

The following equation represents the relationship of the above quantities: 

A = BCD 
This equation may be solved for C or B as follows : 

C- - B- A 
~ BD' ~ CD 

By assuming the various constants of the disk, the apertures may be con- 
veniently laid out in accordance with any desired scheme by following the above 
formula so as completely to scan the image any desired number of times for each 
revolution of the disk. 

1,825,487. Scanning Device. A. O. TATE. Sept. 29, 1931. An endless belt 
is provided with a staggered series of apertures. The belt is looped around a 
multiplicity of guide drums and is driven by rollers at opposite ends of a frame 
structure, so that the apertures are moved successively across the field of a lens 
system for scanning an object within the field of the lens. The object is scanned in 
lines from bottom to top or top to bottom. The band is approximately 160 inches 
in length and has approximately 80 apertures therein, each spaced from the ad- 
jacent aperture at a distance of 4 inches. 

1,825,497. Light Projection Display Apparatus. T. WILFRED. Sept. 29, 
1931. A poly sided screen consisting of a plurality of upright differently faced 
concave sides meeting in thin edges is provided for a display surface. There are 
light projecting means spaced outwardly in front of each of the concave sides, 
the several projecting means being adapted to project cooperatively upon the 
respective adjacent concave sides to produce an ornamental light display for at- 
tracting the attention of a spectator. The projection apparatus is used hi various 
forms of floodlighting architectural displays. 

1,825,598. Process for Producing Combined Sound and Picture Films. 
H. VOGT, J. MASSOLLE, AND J. ENGL. Assignors by mesne assignments to 
American Tri-Ergon Corp. Sept. 29, 1931. The sound and picture records are 
photographed on separate film strips to form separate negatives. The negative 
picture record is photographed upon a portion of a sensitized film not exposed to 
the sound record. The negative sound record is photographed on the same face 
of the sensitized film but on a portion thereof not exposed to the picture record. 
By the separation of the sound record from the picture record, a film record com- 
bining both of these records can be produced without subjecting either record to 
conditions of overexposure or underexposure. 

(Abstracts compiled by John B. Brady, Patent Attorney, Washington, D. C.) 



SOCIETY OF MOTION PICTURE 
ENGINEERS 

OFFICERS 
1931-1932 

President 

A. N. GOLDSMITH, Radio Corporation of America, New York, N. Y. 

Past-President 
J. I. CRABTREE, Eastman Kodak Company, Rochester, N. Y. 

Vice-Presidents 

W. C. HUBBARD, General Electric Vapor Lamp Co., Hoboken, N. J. 
E. I. SPONABLE, Fox Film Corp., New York, N. Y. 

Secretary 
J. H. KURLANDER, Westinghouse Lamp Co., Bloomfield, N. J. 

Treasurer 
H. T. COWLING, Eastman Teaching Films, Inc., Rochester, N. Y. 

Board of Governors 

F. C. BADGLEY, Canadian Government Motion Picture Bureau, Ottawa, Canada 
H. T. COWLING, Eastman Teaching Films, Inc., 343 State St., Rochester, N. Y. 
J. I. CRABTREE, Research Laboratories, Eastman Kodak Co., Rochester, N. Y. 
P. H. EVANS, Warner Bros. Pictures, Inc., 1277 E. 14th St., Brooklyn, N. Y. 
O. M. GLUNT, Bell Telephone Laboratories, New York, N. Y. 
A. N. GOLDSMITH, Radio Corporation of America, 570 Lexington Ave., New 

York, N. Y. 

W. C. HUBBARD, General Electric Vapor Lamp Co., Hoboken, N. J. 
R. F. MITCHELL, Bell & Howell Co., 1801 Larchmont Ave., Chicago, 111. 
J. H. KURLANDER, Westinghouse Lamp Co. Bloomfield, N. J. 
W. C. KUNZMANN, National Carbon Co., Cleveland, Ohio 

D. MACKENZIE, Electrical Research Products, Inc., 7046 Hollywood Blvd., 

Los Angeles, Calif. 
L. C. PORTER, General Electric Co., Nela Park, Cleveland, Ohio 

E. I. SPONABLB, 277 Park Ave., New York, N. Y. 
130 



COMMITTEES 



131 



COMMITTEES 
1931-1932 

( The completed list of committees -will be published in a later issue} 

Color 

R. M. EVANS, Vice-Chairman 

P. D. BREWSTER W. T. CRESPINEL N. M. LA PORTE 

J. L. CASS H. B. TUTTLE 



W. C. HUBBARD 



Convention 
W. C. KUNZMANN, Chairman 



M. W. PALMER 



H. T. COWLING 
W. B. COOK 



Finance 

L. A. JONES, Chairman 
J.I. CRABTREE 
W. C. HUBBARD 



J. H. KURLANDER 

L. C. PORTER 



J. R. CAMERON 

B. W. DEPUE 

C. D. ELMS 



Historical 
C. L. GREGORY, Chairman 

Membership and Subscription 

H. T. COWLING, Chairman 
W. H. CARSON, Vice-Chairman 

R. EVANS 

J. G. T. GILMOUR 



J. KLENKE 
E. E. LAMB 
E. C. SCHMTTZ 



J. A. BALL 
C. DREHER 
P. H. EVANS 
A. C. HARDY 



Papers 

O. M. GLUNT, Chairman 
N. M. LA PORTE 
G. E. MATTHEWS 
P. A. McGuiRE* 
D. McNicoL 



P. MOLE 
K. F. MORGAN 
C. N. REIFSTECK 
T. E. SHEA 



J. O. BAKER 
T. BARROWS 
W. H. BELTZ 
G. C. EDWARDS 
S. GLAUBER 



Projection Practice 
H. RUBIN, Chairman 
J. H. GOLDBERG 
C. GREENE 
H. GRIFFIN 
J. HOPKINS 
R. H. MCCULLOUGH 
P. A. McGuiRE 



R. MlEHLING 

F. H. RICHARDSON 
M. RUBEN 
P. T. SHERIDAN 
L. M. TOWNSEND 



132 



COMMITTEES 



J. L. CASS 
H. GRIFFIN 



Projection Screens 
S. K. WOLF, Chairman 

J. H. KURLANDER 

W. F. LITTLE 



A. L. RAVEN 
H. RUBIN 



F. C. BADGLEY 
B. W. DEPUE 



Publicity 

W. WHITMORE, Chairman 
D. E. HYNDMAN 
F. S. IRBY 



G. E. MATTHEWS 
D. McNicoL 



M. C. BATSEL 
P. H. EVANS 
N. M. LA PORTE 



Sound 

H. B. SANTEE, Chairman 
E. W. KELLOGG 
C. L. LOOTENS 
W. C. MILLER 



H. C. SILENT 
R. V. TERRY 
S. K. WOLF 



L. E. CLARK 
L. DE FOREST 
J. A. DUBRAY 
P. H. EVANS 
R. E. FARNHAM 
H. GRIFFIN 



Standards and Nomenclature 
M. C. BATSEL, Chairman 
A. C. HARDY 

R. C. HUBBARD 

L. A. JONES 
N. M. LA PORTE 
D. MACKENZIE 
G. A. MITCHELL 
G. F. RACKETT 



W. B. RAYTON 
C. N. REIFSTECK 
V. B. SEASE 
T. E. SHEA 
J. L. SPENCE 
E. I. SPONABLE 



R. S. BURNAP 
W. H. CARSON 



Ways and Means 
D. McNicoL, Chairman 
H. GRIFFIN 
F. S. IRBY 



J. H. KURLANDER 
J. A. NORLING 



Chicago Section 

R. F. MITCHELL, Chairman R. P. BURNS, Manager 

B. W. DEPUE, Sec.-Treas. O. B. DEPUE, Manager 

New York Section 

P. H. EVANS, Chairman M. C. BATSEL, Manager 

D. E. HYNDMAN, Sec.-Treas. J. L. SPENCE, Manager 

Pacific Coast Section 

D. MACKENZIE, Chairman G. A. MITCHELL, Manager 

E. HUSE, Secretary H. C. SILENT, Manager 

L. E. CLARK, Treasurer 



CONTRIBUTORS TO THIS ISSUE 

Burke, B. S.: E.E., University of Michigan, 1923; engineer, Westinghouse 
Electric & Manufacturing Company, 1923-27; theater switchboard engineer, 
Westinghouse Electric & Manufacturing Company, 1927-29; developed ther- 
mionic tube control for theater lighting, Westinghouse Electric & Manufacturing 
Company, 1929 to date. 

Curtis, A. M.: Born June 4, 1890, at Brooklyn, New York. United Wireless 
Company, 1907-10; radio engineer, Lloyd Brasiliero S. S. Company and 
Brazilian Department of Agriculture, 1910-13; engineer, Western Electric 
Company, 1917-19; engineer, research laboratory, Western Electric Company 
and Bell Telephone Laboratories, 1919 to date. 

Jones, L. A.: See May, 1931, issue of JOURNAL. 

Rumpel, C. H.: Born 1903 at Kitchener, Ontario, Canada. Graduate, 
Massachusetts Institute of Technology; technical staff, Bell Telephone Labora- 
tories, 1929 to date. 

Shea, T. E.: See March, 1931, issue of JOURNAL. 



133 



SOCIETY ANNOUNCEMENTS 



BOARD OF GOVERNORS 

At a meeting of the Board of Governors held at the Waldorf- 
Astoria Hotel, New York, N. Y., on December 10th, consideration was 
given to the establishment of new committees which would expand to 
a great extent the scope of activities of the Society in directions which 
have so far not been adequately investigated. Among the new com- 
mittees considered was one to deal with non-theatrical home equip- 
ment, a committee on the development and care of film, a Museum 
Committee whose duty it will be to gather historical pieces of motion 
picture apparatus for purposes of exhibition in suitable depositories, 
and a committee on the preservation of film. 

It was decided that the S. M. P. E. Fellowship, created through the 
generosity of Mr. George Eastman, is to be established at the Uni- 
versity of Rochester. Its administration is to be left to the Projection 
Theory Committee, the object of the Fellowship being to conduct 
investigations on problems particularly concerned with or cognate to 
the motion picture art. 

Much discussion was held concerning the financial operations of the 
Society for the fiscal year and on the general matters of entrance fees, 
dues, and subscription rates. 

SPRING, 1932, CONVENTION 

The Board of Governors decided that the names of the cities, New 
York, N. Y., and Washington, D. C., be placed upon the ballot which 
is to be mailed to the entire membership for voting upo tnhe location 
of the Spring, 1932, Meeting. 

NEW YORK SECTION 

A meeting of this section was held on Wednesday, December 9th, 
in the auditorium of the Engineering Societies Building, 33 West 
39th Street, New York, N. Y. Mr. H. A. Frederick of the Bell 
Telephone Laboratories repeated his paper entitled, "Vertical Sound 
Records; Recent Technical Advances in Mechanical Records on 
134 



SOCIETY ANNOUNCEMENTS 135 

Wax," which was presented at the Swampscott Convention on Octo- 
ber 7th. The demonstration which accompanied the paper included 
considerably more elaborate apparatus than that which was used at 
the Swampscott Meeting. Following Mr. Frederick's presentation 
Mr. Leopold Stokowski, director of the Philadelphia Orchestra, ad- 
dressed the meeting, presenting from a musician's standpoint some 
of the problems of recording. The meeting created considerable 
interest, more than seven hundred and fifty people attending in spite 
of the inclement weather. 

The next meeting of the Section is scheduled to be held about the 
second week in January. Announcements will be mailed to all mem- 
bers enrolled in the Section. Those whose names are not on the 
mailing list of the Section, but who wish to receive information con- 
cerning the meetings, should communicate with the general office of 
the Society. 

PROJECTION PRACTICE COMMITTEE 

At a meeting of the Projection Practice Committee held on Novem- 
ber 24th, a program outlining the work to be conducted by the Com- 
mittee during the current year was formulated, being particularly 
directed toward the recommendation of standards of tolerances and 
clearances of projector and sound parts, and the determination of the 
degree of wear of projector and sound equipment which can be al- 
lowed without impairing the quality of the projected picture or dam- 
aging the film. A study of the methods of so-called processing, or the 
treating of finished positive prints to prevent damage during the first 
showing of the film is to be included in the work of the year. It is 
felt that there is a need for more perfect methods which will com- 
pletely eliminate the shedding of emulsion or the accumulation of oil 
and wax in the projector, due to the film. 

At a second meeting of the Committee on Tuesday, December 15th, 
further discussion of the problems of tolerances and clearances in 
projector and sound parts was held, particular attention being paid 
to the points at which tension of the film and wearing of the parts 
occur. The problem of the specifications desirable for projector 
apertures was also discussed at some length, and the work on this 
problem, although not completed, is recommended for the study of the 
Standards and Nomenclature Committee of the Society. 



136 SOCIETY ANNOUNCEMENTS [J. S. M. P. E. 

MEMBERSHIP CERTIFICATE 

Associate members of the Society may obtain the membership 
certificate illustrated below by forwarding a request for the same to 
the General Office of the Society at 33 W. 42nd St., New York, N. Y., 
accompanied by a remittance of one dollar. 



Sonets/Motion Picture Engineers 



INCOSPOOATEO 




Society of Motion Picture Engineers 




LAPEL BUTTONS 






There is mailed to each newly elected member, upon his first 
payment of dues, a gold membership button which only members 
of the Society are entitled to wear. This button is shown twice 
actual diameter in the illustration. The letters are of gold on a 
white background. Replacements of this button may be obtained 
from the General Office of the Society at a charge of one dollar. 



Jan., HK32J 



SOCIETY ANNOUNCEMENTS 
JOURNAL BINDERS 



137 



The binder shown in the accompanying illustration serves as a 
temporary transfer binder or as a permanent cover for a complete 
year's supply of JOURNALS. It is made of black crush fabrikoid, 
with lettering in gold. The binder is so constructed that each in- 
dividual copy of the JOURNAL will lie flat as its pages are turned. 
The separate copies are held rigidly in place but may be removed or 
replaced at will in a few seconds. 




These binders may be obtained by sending your order to the 
General Office of the Society, 33 West 42nd Street, New York, N. Y., 
accompanied by a remittance of two dollars. Your name and the 
volume number of the JOURNAL may be lettered in gold on embossed 
bars provided for the purpose at a charge of fifty cents each. 



SUSTAINING MEMBERS 

Bell & Howell Co. 
Carrier Engineering Corp. 
Case Research Laboratory 

Eastman Kodak Co. 

Electrical Research Products, Inc. 

National Carbon Co. 

RCA Photophone, Inc. 



BACK NUMBERS OF THE TRANSACTIONS AND JOURNALS 

Prior to January, 1930, the Transactions of the Society were published quar- 
terly. A limited number of these Transactions are still available and will be 
sold at the prices listed below. Those who wish to avail themselves of the op- 
portunity of acquiring these back numbers should do so quickly, as the supply 
will soon be exhausted, especially of the earlier numbers. It will be impossible 
to secure them later on as they will not be reprinted. The cost of all the available 
Transactions totals $46.25. 



No. 


Price 




No. 


Price 


1917 { I 


$0.25 
0.25 


1924- 


r !8 
19 


$2.00 
1.25 


1918 7 


0.25 




[20 


1.25 


1920 ft 
1921 11 


1.00 
1.00 
1.00 
1.00 


1925 


21 
22 
23 
24 


1.25 
1.25 
1.25 
.25 


1922 \\ 
1923 ^ 


1.00 
1.00 
2.00 
2.00 


1926- 


25 
26 
27 
28 


.25 
.25 
.25 
.25 



1927 



1928 



1929 



No. 

29 

30 
31 
32 
33 
34 
35 
30 
37 
38 



Price 
$1.25 
.25 

.25 
,25 

.50 



1 

1 

1, 

2. 

2.50 

2.50 

2.50 

3.00 

3.00 



Beginning with the January, 1930, issue, the JOURNAL of the Society has been 
issued monthly, in two volumes per year, of six issues each. Back numbers of all 
issues are available at the price of $1.50 each, a complete yearly issue totalling 
$18.00. Single copies of the current issue may be obtained for $1.50 each. 
Orders for back numlu-rs of Transactions and JOURNALS should be placed through 
the General Office of the Society, 33 West 42nd Street, New York, N. Y., and 
should be accompanied by check or money-order. 
138 



JOURNAL 

OF THE SOCIETY OF 

MOTION PICTURE ENGINEERS 

SYLVAN HARRIS, EDITOR 
Volume XVIII FEBRUARY, 1932 Number 2 

CONTENTS 

Vertical Sound Records: Recent Fundamental Advances in 

Mechanical Records on "Wax" H. A. FREDERICK 141 

Sound Recording From the Musician's Point of View 

LEOPOLD STOKOWSKI 164 
On the Assignment of Printing Exposure by Measurement of 

Negative Characteristics CLIFTON TUTTLE 172 

Utilization of Desirable Seating Areas in Relation to Screen 

Shapes and Sizes and Theater Floor Inclinations 

BEN SCHLANGER 189 

A Method of Measuring Directly the Distortion in Audio 

Frequency Amplifier Systems W. N. TUTTLE 199 

Directional Effects in Continuous Film Processing 

J. CRABTREE 207 

Resume of the Proceedings of the Dresden International 

Photographic Congress S. E. SHEPPARD 232 

Committee Activities: 

Report of the Projection Screens Committee 242 

Abstracts , 255 

Patent Abstracts 258 

Book Reviews 266 

Officers 267 

Committees 268 

Contributors to This Issue 271 

Society Announcements 272 



JOURNAL 

OF THE SOCIETY OF 

MOTION PICTURE ENGINEERS 

SYLVAN HARRIS, EDITOR 



Published monthly at Easton, Pa., by the Society of Motion Picture Engineers. 

Publication Office, 20th & Northampton Sts., Easton, Pa. 
General and Editorial Office, 33 West 42nd St., New York, N. Y. 



Copyrighted, 1932, by the Society of Motion Picture Engineers, Inc. 



Subscription to non-members, $12.00 per annum; to members, $9.00 per annum, 
included in their annual membership dues; single copies, $1.50. A discount 
on subscriptions or single copies of 15 per cent is allowed to accredited agencies. 
Order from the Society of Motion Picture Engineers, Inc., 20th and Northampton 
Sts., Easton, Pa., or 33 W. 42nd St., New York, N. Y. 

Papers appearing in this Journal. may be reprinted, abstracted, or abridged 
provided credit is given to the Journal of the Society of Motion Picture Engineers 
and to the author, or authors, of the papers in question. 

The Society is not responsible for statements made by authors. 

Entered as second class matter January 15, 1930, at the Post Office at Easton, 
Pa., under the Act of March 3, 1879. 



VERTICAL SOUND RECORDS 



RECENT FUNDAMENTAL ADVANCES IN MECHANICAL RECORDS ON 

"WAX"* 



H. A. FREDERICK** 



Summary. This paper describes recent progresswhich has been made in laboratory 
studies of mechanical records of sound cut on a wax disk. Both theoretical and 
experimental investigations indicate that a phonograph record, cut with vertical undula- 
tions instead of the more usual lateral undulations possesses fundamental advantages. 
The principal improvement comes from a marked increase in the volume and frequency 
range over which faithful reproduction may be obtained. A higher volume level can be 
recorded for the same groove spacing and speed. More playing time can be provided 
with a given size of record and volume level since, for these conditions, both the groove 
spacing and speed may be reduced. Improvements in methods of processing the 
stampers and in the record material give a large reduction in surface noise and hence a 
corresponding increase in the volume range. With these improvements the frequency 
range which can be reproduced satisfactorily can be extended nearly an octave to 
8000 to 10,000 cycles. Other improvements incidental to the improvements noted 
above are great improvement in the quality of reproduction obtainable directly from 
a soft "wax" record and a great extension in the life of the hard record. 

At the convention of this Society held at Lake Placid in the fall of 
1928, data were presented showing that a very good frequency charac- 
teristic could be obtained in recording and reproducing by means of 
the "lateral" disk recording system. 1 The data presented at that 
time had to do chiefly with the response-frequency characteristics of 
the elements which entered into that system. The information then 
available, however, about non-linear distortion was somewhat 
limited. That discussion, in addition, did not attempt to cover the 
limitations imposed by background noise commonly called "surface" 
or "needle scratch." 

In most commercial uses of lateral records, surface noise has 
imposed very serious limitations. In many cases this noise has been 

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. 
(Repeated at a meeting of the New York Section, December 9, 1931, at New 
York, N. Y.) 

** Bell Telephone Laboratories, New York, N. Y. 

141 



142 



H. A. FREDERICK 



[J. S. M. P. E. 



suppressed by the use of so-called "scratch" filters. These have 
effectively quieted the reproduction but only by the sacrifice of an 
important portion of the recorded band of frequencies which are 
above 3000 to 4000 cycles. Investigations have been carried on to 
determine the fundamental causes and the characteristics of the 
surface noise in order that, with a better understanding, it might be 
more effectively reduced and without such a sacrifice. 

In addition to the limitations imposed by surface noise, other 
studies have indicated that, with the available reproducers for lateral 
cut records, the needle point may fail to follow the center of the groove 
accurately when the curvature becomes too sharp, and may skid from 
side to side by varying amounts, depending on the record and the 
characteristics of the reproducer being used. Studies have proceeded 



RECORDER 
STYLUS 




A = POINT OF CONTACT OF REPRODUCER 
STYLUS WITH GROOVE 

FIG. 1. Distortion in a lateral groove. . 

relating to the physical characteristics necessary in a reproducer in 
order that it may faithfully follow a groove. These studies have led 
us to expect superior performance from a groove cut with vertical un- 
dulations than from one with lateral undulations. These records are 
similar in principle to those used by Mr. Edison. With the lateral 
groove there is distortion due to the fact that the sound is recorded 
with a chisel-shaped stylus and reproduced with a round stylus; 
also that in reproduction the bearing point of the stylus against the 
groove shifts forward and backward as the needle rounds a curve. 
These effects are illustrated in Fig. 1. With vertical records the first 
of these effects, sometimes called the "pinch" effect, is absent, but a 
shifting of the bearing point of the reproducing stylus forward and 
backward occurs if a round stylus is used. It is doubtful if a chisel- 
shaped reproducing stylus or a stylus with an elliptical point can be 
justified due to the increased cost and complication, and in considera- 



Feb., 1932] 



VERTICAL SOUND RECORDS 



143 



tion of the rather small amount of distortion which this would 
eliminate. Some qualitative idea of what takes place with vertical 
undulations may be gained from Fig. 2, in which a sine wave is shown 
together with the resulting positions of the stylus point. For a given 
stylus tip radius and for a given recording level this effect increases 
with frequency. 

This failure of a stylus point to follow a vertical record with great 
accuracy is, of course, due to the finite length of the stylus point 
along the groove. A fact which relieves this situation is that speech 
and music and most other sounds which we are interested in re- 
cording contain much less energy in the high than in the low fre- 
quency range. 2 

Frequency analyses of surface noise have been made using a variety 



RECORDER 
STYLUS 



REPRODUCER STYLUS POSITIONS 




FIG. 2. Distortion in a vertical groove. 

of reproducers and record materials. In general, these frequency 
characteristics have been found to be very largely influenced by the 
characteristics of the reproducers, but do not show any marked differ- 
ences as between lateral and vertical recordings. Frequency charts of 
surface noise taken with a vertical reproducer having a very flat 
frequency characteristic over the audible range have shown the 
surface noise to be relatively richer in high frequencies. The distribu- 
tion of surface noise energy below 10,000 cycles from a cellulose 
acetate pressing is shown in Fig. 3. The amount of recorded sound 
energy in the low frequency range, i. e., below about 2000 or 3000 
cycles, however, is large relative to that in the higher frequency 
ranges. Moreover, the characteristics of many lateral reproducers 
have been such as to accentuate surface noise between 3000 and 
5000 cycles. Hence the use of "scratch" filters for the elimination of 



144 



H. A. FREDERICK 



[J. S. M. P. E. 



the high frequency components of the surface have made a large effec- 
tive reduction in noise without any material loss in loudness of the 
sounds of interest. The loss in loudness at the higher frequencies 
has also reduced the audible distortion due to poor traction and, 
although the loss of the higher frequencies is serious, it has been 
held by many that the end has justified the means. Surface noise 
is probably caused by a more or less random distribution of im- 
pulsive shocks on the needle due to minute irregularities in the 
record. It has been common practice in lateral recording to use 
record material containing a certain amount of abrasive in order 
to grind the needle to fit the groove. The irregularities due to 
the abrasive would logically be expected to produce a scratchy 
noise of much the character with which we are all familiar. A 
5000-cycle note of the same loudness as a 10,000-cycle band of 
surface noise using a reproducer with a flat characteristic would have 




40 50 



5OO 1000 

FREQUENCY IN CYCLES PER SECOND 



5000 



O,OOO 



FIG. 3. Energy distribution of surface noise from a cellulose acetate record. 

an amplitude of only about 0.000001 inch. In order to reduce the 
surface noise to the point where it is no longer troublesome, it appears 
necessary to eliminate irregularities at least down to this order of 
magnitude. It has been found that, if the usual abrasive record were 
replaced by an unabrasive record pressed of a very clean homogene- 
ous material such as cellulose acetate, the surface noise caused by the 
record material itself would be greatly reduced. Such a change, 
however, by itself, has been found to give a comparatively minor 
improvement ; for, when this cause is moved well into the background, 
other causes of surface noise of practically the same order of magni- 
tude as that due to the abrasive of a shellac record become controlling. 
The next process which it has been found necessary to improve has 
been that of rendering the surface of the original wax electrically 
conducting. The usual methods of graphiting or brushing with fine 
electrically conducting powders have been found unsatisfactory. 



Feb., 1932] VERTICAL SOUND RECORDS 145 

Recourse has therefore been had to one of the earlier methods used in 
phonograph practice, namely, cathode sputtering 3 of the wax. This 
method was not devoid of difficulty, however. With the best sputter- 
ing technic the usual thick "waxes" are heated to such an extent as 
to injure or destroy their finely engraved surfaces. By using a very 
thin layer of wax flowed on a metal surface it is possible to keep it cool 
during the sputtering operation. It is thus possible to apply an 
extremely uniform, smooth, and tenacious surface of metal of adequate 
thickness in a very few minutes. This can be electroplated by the 
ordinary methods, and the electroplate used for pressing the final 
record. By using this thinly flowed wax, it is possible to obtain 
a surface texture which is extremely smooth and homogeneous and 
which is also free from the mechanical strains incident to shaving 
the waxes by the methods previously commonly used. In addition, 
waxes of this type possess obvious advantages in ease of transporta- 
tion, ruggedness, etc. When the noise due to the two causes dis- 
cussed above has been removed or largely reduced, a third source 
of noise is apt to become prominent. This involves the reaction 
of the wax shaving on the recording stylus, which appears on the 
final record as "clicks" when the shaving breaks or is removed 
in a non-uniform manner. It has, however, been found possible 
by suitable design to provide a recorder, stylus, and suction arrange- 
ment such that the shaving is removed in a very smooth stream, thus 
eliminating this type of noise to a large extent. 

It has been common practice in the past to provide duplicate 
stampers by electroplating the first stamper or "master" to obtain a 
negative metal record. This in turn has been plated to provide the 
duplicate stamper. A convenient and quick alternative method is 
provided by sputtering and plating a suitable pressing made directly 
from the "master." 

These improvements in the methods of engraving and processing, 
and in the final record material are more or less applicable to either 
type of recording, lateral or vertical. Their full value, however, can 
only be realized provided full advantage may be taken of the increased 
frequency range which greater quietness permits. It is possible to 
take advantage of this improvement to effect other improvements or 
economies rather than to use it all in the one direction of decreased 
noise. In amount, the reduction in surface noise from that of present 
commercial records will differ depending on the frequency range 
reproduced. 



146 H. A. FREDERICK [J. S. M. P. E. 

If a blank groove record, made with the improvements noted above, 
is reproduced by a reproducer which is uniformly responsive up to 
10,000 cycles, the surface noise is 20 db. below that of an old type 
record reproduced in the same manner. If, however, all frequencies 
above 5000 cycles are eliminated in each case, the difference is 15 db. 
If now the noise of the new record reproduced to 10,000 cycles is 
compared with the old record reproduced to 5000 cycles only, which 
is the comparison of greatest practical interest, the difference in noise 
is about 15 db. In addition, it is possible to take advantage of the 
fact that most sounds to be recorded contain less energy in the high 
frequency range than in the medium or low frequency range, and to 
record the higher frequencies at a level somewhat higher than normal. 
In reproduction these higher frequencies are then correspondingly 
reduced by the reproducing amplifier or circuit. It is thus found 
that a further reduction of about 10 db. in surface noise can be ob- 
tained, the amount depending somewhat on the high frequency cut-off 
of the reproducer or circuit. This effect occurs chiefly between 5000 
and 10,000 cycles. 

The ''volume range" for any particular frequency band is usually 
considered to be the difference in decibels between the loudness of the 
surface noise and the loudness of the maximum recorded sound which 
the record can accommodate when reproduced faithfully over this 
frequency range. With the lateral records of the past, reproduced to 
5000 cycles, this volume range may be stated as about 25 to 30 db. 
This figure obviously will differ somewhat for different cases, depend- 
ing on the character of the sounds recorded and on the degree of 
excellence obtained with the recording and processing methods 
throughout. With vertical recording the reductions in surface noise 
described above increase the volume range for a 5000-cycle band of 
frequencies to from 50 to 55 db. For 10,000-cycle reproduction the 
volume range is 45 to 50 db. Obviously, these new facilities open 
the door for very great improvements in fidelity of reproduction and 
for the reproduction of many effects not possible in the past. In 
many cases it means that the surface noise may be reduced to in- 
audibility. 

Lateral records have usually been cut with a stylus having a tip 
radius between 0.002 inch and 0.003 inch. The angle between the 
two sides has, in this country, commonly been about 90 degrees. The 
groove has been 0.002 inch to 0.003 inch deep and about 0.006 inch to 
0.007 inch wide. The groove spacing has been 0.010 inch to 0.011 



Feb., 1932] VERTICAL SOUND RECORDS 147 

inch so that the uncut space between blank grooves has been 0.003 
inch to 0.004 inch. If one groove is not cut over into the next, the 
maximum amplitude which can be used is limited to about 0.002 inch. 
If the usual loudness of the record is to be maintained it is necessary 
to maintain this spacing between grooves. 

With vertical records it has been found desirable, particularly 
where a very loud record is to be made, to use a recording stylus with 
approximately the same tip radius as previously used with lateral 
records, but to reduce the divergence between the sides of the stylus 
above the tip. In addition, it has not been found necessary to 
provide any clearance space between grooves. In fact, it has been 
found entirely satisfactory to have the side of one groove cut con- 
sistently into the next. It is therefore entirely feasible to increase 
the number of grooves per inch from the usual 98 to between 125 and 
150, at the same time that the recording level is increased. When 
using this recording stylus with the lesser divergence for cutting a 
record with 125 to 150 threads per inch, it has been found desirable to 
make the groove about 0.007 inch wide and about 0.003 inch deep. 
The maximum amplitude may, under these conditions, be increased 
about 4 db. It has been found possible, however, to obtain satis- 
factory results with most waxes even though the normal depth of the 
groove is increased to as much as 0.004 inch to 0.006 inch. In this 
case, the recorded level may be increased 6 db. This increase in the 
recording level obviously increases the volume range by a like amount. 
If occasionally, due to a loud crash of sound, the recording stylus 
completely leaves the wax, the reproducer will still "track" satis- 
factorily; that is, continue in the correct groove. The corresponding 
situation with a lateral record where one groove cuts into another is, 
of course, fatal since in such a case the reproducer will usually cross 
into the next groove. It has been found desirable with vertically cut 
records to use a permanent reproducing stylus in order to reduce the 
vibrating mass of the reproducer to a satisfactory value. This 
stylus point remains sharp in contrast with the old steel needles used 
with lateral records, and therefore will reproduce satisfactorily 
undulations of sharper curvature. In other words, for the same 
amplitudes the linear speed of the record may be reduced. Practically, 
it may be undesirable to reduce or change the rate of rotation of a 
record from the commercially used value. It is, however, feasible, 
to decrease the internal groove diameter recorded on the 33 rpm. 
record to about 6 inches for a 10,000-cycle frequency range. By 



148 



H. A. FREDERICK 



[J. S. M. p. E. 



the combination of the various elements mentioned above, it is 
feasible to record for 15 to 20 minutes on a 12-inch record and for 
10 to 12 minutes on a 10-inch record. This involves the use of 
about 200 grooves per inch and a decrease in the recorded level 
to about the level of laterally recorded records using 98 grooves 
per inch. Of course, longer recordings can be made in the same space 
if the recorded level is decreased (more grooves per inch), or if the 
upper frequency cut-off is decreased (decreased rpm. and inner 
diameter). However, these changes may introduce tracking difficul- 
ties if carried too far and must be well justified by other considerations. 
Laterally and vertically cut records drive the reproducer point 
quite differently. Laterally cut records drive the point from both 
sides but the point rarely follows the center of the groove with great 
exactitude. It deviates from the center by amounts chiefly depen- 



-50 
-60 
-70 

-80 
4 












db= 1 VOLT ACROSS LOAD IMPEDANCE OF 1 OHM FOR 
VIBRATORY VELOCITY = 1 CENTIMETER PER SECOND 






































































































*" 


























































































50 100 500 1000 5000 IO,C 



FREQUENCY IN CYCLES PER SECOND 

FIG. 4. Response frequency characteristic of an experimental vertical 
reproducer driven by cellulose acetate records. 

dent upon the mechanical impedance of the reproducer. A vertically 
cut record, on the other hand, drives in only one direction. The 
restoring force is due . chiefly to the elasticity of the supporting 
structure of the reproducer, the normal restoring force being equal 
to the total weight on the needle minus the weight of the moving or 
vibrating part. The stylus point will always remain in contact with 
the record unless the forces set up by the undulations exceed this 
normal restoring force. Operation should always be below this 
limiting condition. This sets definite requirements on the mechanical 
impedance of the vibrating parts and, unless this condition can be 
met, reproduction of extreme frequencies by vertical records is im- 
possible. With the vertical reproducers which we have used, the 
stylus can follow sudden downward motions of the record groove even 
to accelerations about a thousand times that due to gravity. With 
laterally cut records, there is no definite limiting condition analogous 



Feb., 1932] 



VERTICAL SOUND RECORDS 



149 



to the above. However, it appears easier in practical design to 
reduce greatly the mechanical impedance of vertical than of lateral 
reproducers. Practical experience has shown that the mass can be 
so reduced as to reproduce up to well above 10,000 cycles and the 
stiffness reduced so as to reproduce down to the order of 20 cycles. 
In fact, there appears to be considerable margin on this score. This 
makes it possible to reduce the weight with which the reproducer 
point bears on the record to between 2 and 20 per cent of what has been 
used with most commercial lateral reproducers. This reduction in 
stylus or needle point pressure has been found to decrease the wear 
on the record very greatly, with the result that its life has been 
considerably increased. Tests have shown that the first few thousand 
playings cause negligible deterioration, and even several hundred 
thousand playings do not show excessive wear if the record is properly 
protected from dust and dirt. 



RESPONSE IN DECIBELS 

fc. .C 














Odb = l CENTIMETER PER SECOND PER VOLT PER OHM 


























x-^ 






























\ 
\ 
































^ 


** 


,^-- 






















































































50 100 500 1000 5000 10,0 



FREQUENCY IN CYCLES PER SECOND 

FIG. 5. Response frequency characteristic of an experimental vertical recorder. 

A highly satisfactory method of providing a reproducer for 
vertically cut records has been to use the type of structure with which 
we are all familiar in loud speaker design ; namely, that in which a coil 
moves in a radial magnetic field. Such a reproducer is simple and 
sturdy. Its performance is linear over a wide amplitude range; it 
may be made extremely light and, at the same time, is quite efficient. 
The coils used have had a diameter of between 0.1 and 0.2 inch, and 
the total mass of the vibrating system, including the diamond or 
sapphire stylus, has varied with different models from 5 to 35 milli- 
grams. The total force on the record has been reduced from about 
150 grams to between 5 and 25 grams, the lighter structure being 
used when playing from a soft wax. With the larger of these designs 
it has been found possible to obtain efficiencies which are comparable 
with the efficiency of the Western Electric oil-damped reproducer used 
with lateral records. No difficulty has been experienced due to 



150 



H. A. FREDERICK 



[J. S. M. P. E. 



failure to follow the groove if the reproducer is mounted on a simple 
pivoted arm, as in the case of lateral reproducers. Due to their very 
small mass they operate quite satisfactorily even though the record 
turntable fails to operate in a true plane, and even though the record 
be considerably warped. 

The response of the moving coil vertical reproducer is practically 
constant over a very broad frequency range. It is shown in Fig. 4, 
which is the characteristic of an experimental model taken with 
cellulose acetate pressings. 

The design of a recorder for use with vertically cut records involves 
no fundamentally new problems over those used with laterally cut 
records which have been described previously. 4 It is still desirable 
to design the recorder to approximate a constant amplitude character- 
istic for the lower frequency range and a constant velocity character- 



RELATIVE RESPONSE IN d 
P ri, 1 _ rv 

^ ,uO C 
























































. 




~ 






=. 


^ 


7 = 


-tf^w. 


-* - 


^^1 


r= 


^ 


= 


Ml 


_ 


_- 


i ^ 


- 


^=* 


^a 


r^ 


^ 


\ 




















































50 100 500 1000 5000 10,000 
FREQUENCY IN CYCLES PER SECOND 

. Over-all response frequency characteristic (recorder + reproducer + 



network + amplifier). 

istic for the higher range. This frequency characteristic has been 
often shown and is familiar to all. The same recorders which have 
been used for lateral recording can usually be converted for vertical 
recording by the addition of a comparatively simple link system and 
are quite satisfactory if a high frequency cut-off of 6500 to 7000 
cycles is acceptable. It is, however, desirable to have a higher high- 
frequency cut-off. Such a recorder has been used in making many of 
the records which we have studied. Its frequency characteristic is 
shown in Fig. 5. 

The response of the oil-damped lateral reproducer is greatest at 
the very low frequencies. Its response decreases with frequency, 
this decrease in the lower frequency range compensating more or less 
for the increase of response of the recorder with frequency. Because 
of the flat characteristic of the vertical reproducer, it has been found 
desirable to compensate in the reproducing amplifier or circuit for 



Feb., 1932] 



VERTICAL SOUND RECORDS 



151 



the low response of the vertical recorder at the lower end of the 
frequency scale. A frequency characteristic for the combination of 
recorder, reproducer, amplifier, and network is shown in Fig. 6. 

It has been found with vertical records that speech is reproduced 
with considerably improved naturalness and that the word endings, 
sibilant sounds, etc., are much more distinct. The sounds of the 



RESPONSE IN DECIBELS 
.5 o 6 8 je 
















































^ 


/ S 


V 


/ 


w_^ 


AA 


^o 


r 


^. 


\/ 






^ 


j . 


~x_ 


SJ 


t 


^_> 


/ 




^ 












Odb= SOUND PRESSURE OF 1 BAR AT MEASURING 
POSITION WITH 0.25 WATTS ELECTRICAL INPUT 


























































D 50 100 500 1000 5000 10,0 
FREQUENCY IN CYCLES PER SECOND 

i. 7. Response frequency characteristic of combined high and low 
frequency loud speakers. 



different instruments in an orchestra, particularly when playing a 
loud passage, are reproduced with very great individuality and 
clarity. Results of this kind are difficult to describe and should be 
heard to be appreciated fully. If records such as those described are 
reproduced using various low pass filters, the loss of distinctness due 
to the elimination of frequencies above even 7000 cycles is easily 




30 



50 



5000 10,000 20,000 

FIG. 8. Field calibration of a moving coil microphone. 



500 1000 

FREQUENCY IN CYCLES PER SECOND 



noticeable, whereas little or no difference in needle scratch or surface 
noise may be observed, this being almost wholly absent in all cases. 
The latter statement holds whether the records contain speech or 
music or if blank grooves be reproduced. In listening to such records 
a loud speaker has been used which is essentially flat over a large 
portion of the range of audibility, its characteristic being as shown. 5 



152 H. A. FREDERICK [J. S. M. P. E. 

(Fig. 7.) The reproducer frequency characteristic, as shown in Fig, 4, 
is essentially flat to 10,000 cycles. A corrective network has been used 
which compensates for the low frequency droop in the recorder 
which, at the high frequency end, is, as shown in Fig. 5, essentially flat 
to 9000 cycles. Thin metal-backed waxes have been used which, 
after recording, have been rendered electrically conducting by metal 
sputtering. The moving coil microphone has been used, 6 the charac- 
teristic being as shown in Fig. 8. The records have been pressed of 
cellulose acetate. 

REFERENCES 

1 FREDERICK, H. A.: "Recent Advances in Wax Recording," Trans. Soc. 
Mot. Pict. Eng., 12 (1928), No. 35, p. 709. 

2 FLETCHER, H.: "Speech and Hearing," D. Van Nostrand Co., Inc., New York, 
N. Y. (1929). 

3 GUNTHERSCHULZE, A.: "Cathode Sputtering," Zeit. f. Phys., 36 (Jan. -May, 
1926), p. 563. 

4 MAXFIELD, J. P., AND HARRISON, H. C.: "High Quality Recording and 
Reproducing of Music and Speech," Trans. A. I.E. E., 45 (Feb., 1926), p. 334. 

6 BOSTWICK, L. G.: "An Efficient Loud Speaker at the Higher Audible Fre- 
quencies," /. Acoustical Soc. of Amer., 2 (Oct., 1930), No. 2, p. 242. 

6 JONES, W. C., AND GILES, L. W.: "A Moving Coil Microphone for High 
Quality Sound Reproduction," /. Soc. Mot. Pict. Eng., 17 (Dec., 1931), No. 6, 
p. 977. 

DISCUSSION 

MR. RICHARDSON: What is an L. P. filter? 

MR. FREDERICK: An L. P. or low pass filter is one that cuts out everything 
above the particular frequency noted, and transmits everything below this 
"cut-off" frequency. 

PRESIDENT CRABTREE: I think I have pointed out on several occasions that 
the public has been satisfied to date with the reproduction of speech, but not 
with the reproduction of music, because of its lack of range, both frequency and 
volume. This demonstration has shown that the extent to which the frequency 
range can be covered is excellent. 

I am afraid that the range of volume is still inadequate for providing a facsimile 
of orchestral music. But I think this demonstration shows an epoch-making 
advance in sound reproduction. I don't believe that I have ever heard a re- 
production of a film record that is as satisfying as some of the passages we have 
just heard. 

While I do not predict that the producers will hasten to adopt wax records 
immediately, this performance will make them sit up and get busy, and either 
match this quality on film or turn over to the disk. 

MR. RICKER: We do not get quite the full benefit of these excellent records 
in this room. This room lacks in acoustical qualities for a proper appreciation 
of the magnificent work done. 



Feb., 1932] VERTICAL SOUND RECORDS 153 

MR. PALMER: We are always being told that the radio can produce better 
sound in the home than talking pictures can in the theater. Can Mr. Frederick 
give us any information as to whether the quality 'of the music reproduced here, 
the fidelity of reproduction, is as good as or better than what the best radio 
receiver can furnish? 

MR. FREDERICK: I hesitate to hazard an answer to that, as I am not familiar 
with the characteristics of all radio receivers. The reproduction of many re- 
ceivers that I have heard was greatly inferior in quality to the reproduction to 
which we have just listened, but I prefer to let someone who knows more about 
that particular field attempt to answer your question. 

I think you have been given a very definite picture of what those frequency 
characteristics mean by listening with the different filter settings. That is why 
we played the records with filters in and out so much of the time. One has to 
hear these things again and again, and even then he would have to check up his 
ears every once in a while, in order to have an accurate appreciation of what 
they are hearing. 

MR. MAXFIELD: Was not the reproduction level of the orchestra record as 
reproduced here louder than would be heard in the center of the orchestra seats 
in the theater where it was made? I frequently make tests in that theater, and 
usually sit in the center of the orchestra. My impression here at the back of 
the room is that the reproduction of the loud parts is louder than they appear 
to be in the theater, in fact, a little uncomfortably loud. 

MR. FREDERICK: I believe it was. 

MR. CARLTON: What type of acetate was used for the new record? What 
method was used for the production of the cellulose acetate from which the 
record was made? 

MR. FREDERICK: I cannot tell you in great detail. It was a very pure acetate. 
We obtained it from various sources. Du Pont, for example, has supplied it. 

MR. CARLTON: Is it molded? 

MR. FREDERICK: Yes, with slightly higher temperatures than are used for 
most other record materials. 

MR. HICKMAN: I believe if the Bell Telephone Company were to present this 
entire outfit to an average person living in an apartment, and provide an easy 
means of adjusting the quality of reproduction to suit his taste, you would find 
that in general he would eliminate all components having a frequency greater 
than about thirty-five hundred. If the same person were to listen at a hole in 
the wall leading to an auditorium holding a good orchestra, the hole being dis- 
guised by a loud speaker design, he would tell you that the reproduction was 
rather good but was deficient in low frequency response. As gramophones are 
getting bigger and bigger, if you produced one big enough for a man to get inside, 
and let him speak through a speaker aperture, the observer would tell you that 
it was pretty good but not quite like the human voice. There has grown up, 
since the reproduction of canned music, a sort of new standard of what is desirable. 

Why is it that instinctively we object to what should be the more correct 
form of recording? I am speaking as a layman as an enthusiastic amateur 
musician. 

Is it possible that, when the most perfect reproduction has been made, in 
picking up the sound from the record a high frequency chattering is created 



154 H. A. FREDERICK [J. S. M. P. E. 

which cannot be expressed as harmonics but as a slight disagreeable individuality 
imparted to the record after, say, a frequency of five thousand, which we would 
rather have cut out? 

MR. FREDERICK: I do not think that your question about reproducing these 
high frequencies was directed particularly to me or that you expect me to answer 
about the tastes of people. And I am not sure that I understand your last question. 

PRESIDENT CRABTREE: I think Mr. Hickman wants to know why the fre- 
quencies above five thousand seem to annoy one in the home. 

Recently a friend recommended for my radio a new speaker which had a 
straight curve up to I don't know how many thousand cycles. I obtained this 
speaker and compared it with my old speaker, which certainly does not reproduce 
above five thousand cycles. After repeated tests my observation was*exactly 
the same as Mr. Hickman's that in spite of the high frequencies, I preferred 
the low ones. The frequencies higher than about five thousand apparently 
were annoying, and seemed to irritate the ear. From this observation, it seems 
that straight line reproduction is not always necessary, but depends on the condi- 
tions of the room in which the reproduction occurs and on the tastes of the indi- 
vidual. 

MR. HICKMAN: I am not questioning whether the reproduction shall be linear. 
I am asking why, when apparently the reproduction is most faithful, we do not 
find the reproduction of the high frequencies pleasing. I want to know whether 
some particular form of high frequency distortion is introduced in the pick-up, 
or later on, which is not in the musical record. 

MR. FREDERICK: I do not think that anybody knows enough to really answer 
your question. And I doubt very seriously if a simple answer could be made if 
anyone did know enough. However, this point is extremely important. When 
you extend the high frequency range, if there is any distortion anywhere in the 
system it may be made audible and distinctly annoying, whereas it was previously 
inaudible. If the frequency range is to be extended upward a distinctly higher 
grade of performance must be obtained of all parts of any reproducing system 
than may seem perfectly tolerable with a lower high frequency cut-off. 

The loud speakers so far designed are certainly not perfect. The curves 
shown in the paper indicate that by far the most jagged and roughest curve of 
all was that of the loud speaker. It seems reasonable to think that these irregu- 
larities, which certainly must mean a certain amount of resonance and "hanging 
on" of the sounds, must have an effect on the ear. 

Now, if we could get a perfect loud speaker and remember, that a perfect 
loud speaker means that it must be considered in conjunction with the particular 
room in which it is used the results would undoubtedly be greatly modified. 
The characteristics of a loud speaker will be different in one room from another, 
and may be quite different in different parts of the same room. But placing our- 
selves at a particular place in a particular room, and having an ideal loud speaker 
to project the sound, I personally am convinced that we would, as soon as we were 
used to it, all vote for as broad a range as we could possibly get, and the most 
perfect or straight-line reproduction. 

The trouble is, as we advance in our halting manner, we often make an im- 
provement which shows up defects which were previously inaudible. 

MR. EVANS: Is it not possible that the curves may not tell the complete story? 



Feb., 1932] VERTICAL SOUND RECORDS 155 

The curves that we have seen are for continuous tones single frequencies and 
do not take into account transient effects that may exist. If we knew more 
about transients, might it not be possible to answer the questions that have 
been asked here? 

MR. FREDERICK: I think so. At least, it would take us further. 

MR. MILLS: I speak not only as a layman, as some of the others have spoken, 
but also as one peculiarly inept in music. I have at times attended symphony 
concerts and orchestral renderings, and suffered from the higher and rasping 
violin overtones, and from irritating high frequency sounds of brass instruments 
and cymbals, and it may be that I prefer to listen to a little thirty-five hundred 
cycle cut-off loud speaker, and interpret its output as music. 

But it may be that those who have a wider appreciation of music than I, and 
a greater discernment, would prefer the more nearly complete reproduction 
which includes the higher overtones. 

I should like to ask Mr. Frederick whether he would briefly summarize four 
or five points: What is the increased range of loudness which this new record 
is capable of providing, over and above the previous loudness range? What is 
the increase in frequency range? What is the increase in time recorded under 
normal conditions? And what is the increase or decrease between the cellulose 
acetate record and the normal shellac, in ground noise, at the various frequencies 
which are important? 

MR. FREDERICK: The volume range was stated in the paper as being about 
twenty-five to thirty decibels for most records. I have called them shellac 
records. They are not simply "shellac" records, but shellac plus a lot of other 
technic which accompanies it. The volume range with the type of record demon- 
strated here lies between fifty-five and sixty decibels, according to the best data 
we have. The improvement is not due merely to cellulose acetate. It is due 
to a combination of changes. If one of three or four causes, all approximately 
equal, is eliminated from consideration, the improvement which will have been 
made in the total effect is of course fairly small. Our observations have led us to 
believe that with the old records the noise due to the shellac was somewhat 
greater than the other noises, but only a few decibels greater. As soon as it was 
reduced a little the other components of the noise came into evidence. 

Regarding the time of recording, I tried to summarize this matter in the 
paper, but it is difficult to give any simple and definite figure to cover the entire 
question of playing time. . On the older, lateral record, a greater number of 
grooves per inch was sometimes used. Something has to be sacrificed to do this, 
but it may be worth while. Edison put out hill and dale records which played 
thirty or forty minutes. They were not successful because they did not have 
certain other characteristics which were needed. But as far as playing time is con- 
cerned, that is something on which I don't believe you can make any simple 
statement. 

PRESIDENT; CRABTREE: For a ten-inch record, how long will a lateral play 
and how long a vertical? 

MR. FREDERICK: The usual ten-inch lateral record will play three minutes, 
I believe, and on this type we have thought it good practice to make it play ten 
or twelve minutes.. But if the "game is worth the candle" the time can be made 
longer. You have to sacrifice something else, however, to do so. 



156 H. A. FREDERICK [J. S. M. P. E. 

MR. VICTOR: Is there a relation between frequency and volume? 

Perhaps the high frequencies carry farther, and perhaps it may be possible 
to introduce an automatic modulator of some kind for the home, that might tone 
down the high frequencies to a level that would be more pleasing to the ear. 

MR. FREDERICK: Of course it is the easiest thing in the world to get rid of 
them. The trouble is to get them. High frequencies, I think, are generally 
found not to carry in distance as well as the low frequencies. A good example 
of this is furnished by a man making a speech in an open space. As you walk 
toward him, from a distant point, you first hear the sound of his voice but cannot 
understand a word. As you get closer, you perceive more of the high frequencies, 
until, when you get close enough to get the frequencies on which his articulation 
depends, you can understand what he is saying. This is not simply a matter of 
how the various components are transmitted, but is also concerned with the fact 
that the lower frequencies in speech, as in music, are usually very much stronger. 

PRESIDENT CRABTREE: What Mr. Victor had in mind was a means of con- 
trolling the various frequency components, not to cut them out entirely, but 
selectively to diminish their volume. 

MR. VICTOR: That is right. When a soprano voice comes over my radio I 
usually reduce the volume. 

MR. FREDERICK: If you should go to a concert to hear a first-class soprano, 
you would not think of doing such a thing. If you should go to hear a first-class 
orchestra you would not expect to do it. When we do things of that kind I 
believe we only try to compensate for the faults of the equipment. 

MR. RICHARDSON: When I said that the sound was best, according to my 
judgment, with the seven thousand cycle components included, I did not mean 
that it was most pleasing, but most natural. A railway whistle, of high fre- 
quency, is annoying to everyone. But everyone rather likes to hear a steamboat 
whistle, which is of low frequency. 

Some people enjoy a soprano, but they are the exception. But I do not believe 
that there is anyone in this room who would not enjoy the sound of a contralto 
singing "Silver Threads among the Gold" or something of that kind. The 
sounds are pleasing to everyone. The high frequencies the soprano, the loco- 
motive whistle are annoying to the nerves, not to the ear. 

MR. SHEA: I think there is a great deal in what Mr. Richardson has said. 
It is probably true that most sounds which are startling or grating, as some 
people call them, have a large high frequency content. It seems to be the general 
experience that for the high frequencies you must have clearer reproduction. 

MR. KELLOGG: Quite a little comment seems to have been inspired by the 
idea that during some of the numbers there was a kind of unnaturalness, par- 
ticularly where there was fairly complex music for example, the orchestra. 
It sounded to me as though it might possibly have been due to some imper- 
fection in the correspondence between input and output non-linear distortion 
as we sometimes call it, or it might have been an effect such as one gets when 
in a room with an orchestra and the reverberation is rather high, particularly 
in the high frequency range. I should be interested to know in the case of the 
orchestra recording whether the acoustics of the recording room were such as I 
have described. 

I have another question: The piano was very steady and firm, a condition 



Feb., 1932] VERTICAL SOUND RECORDS 157 

which obtains only when the turntable is rotating very steadily. A year and 
a half ago a turntable mechanism was described wherein a great deal of refinement 
was gone into, to avoid speed irregularities. I should be interested to know 
whether that type of turntable is used both in recording and reproduction in 
this case. 

MR. FREDERICK: In answer to the first question, as to whether there was not 
some non-linear distortion somewhere in the system, I do not think there is any 
question but that there is, but I believe that there was perhaps less of it than we 
sometimes hear. 

The conditions of the pick-up of the sound were not what we should have 
chosen if we had had a place where we could move the microphone about. A 
single microphone had to be placed close to the conductor; those conductors 
to whom I have spoken, and in whose opinions I have great confidence, insist 
that they cannot judge from the conductor's position what the orchestra should 
sound like. They have to permit another conductor to rehearse the orchestra, 
so that they may go out into the body of the hall to get the correct effect. As 
to turntable speed control, the recording machine was the usual Western Electric 
Company recording machine. The reproducing turntable was driven by a 
synchronous motor with multiple belt speed reduction. 

MR. MAURER: How abrupt was the cut-off of the low pass filter used in this 
demonstration ? 

MR. FREDERICK: The cut-off is quite abrupt. A matter of a few hundred 
cycles means a great many "db.'s." 

MR. THOMPSON: Is this type of reproducer more responsive to vibrations of 
the turntable, or irregularities of that kind, than is the present lateral pick-up? 

MR. FREDERICK: Of course these records would not play on a seventy-eight 
rpm. turntable, because they were recorded at thirty-three rpm. They are 
far less sensitive to certain types of irregularity, due to the small mass of the 
pick-up and the small pressure on the record. We had a record one time that 
was pressed with a stamper, that had been bent at least an eighth of an inch out 
of plane, out on one edge, so that there was a big bulge in the record. The re- 
producer tracked over this without difficulty, and no trouble was experienced in 
the reproduction in doing so until the pick-up reached the point where the 
frame of the reproducer hit the bulge in the record. 

As to vibration, I should hesitate to say, because it would seem off-hand that 
a vertical reproducer would tend to be somewhat sensitive to it. So far as my 
experience has shown, it has seemed to be certainly no more sensitive to that 
type of trouble, and my impression is that perhaps it is a bit less sensitive. But 
I should hesitate to make a definite answer on that. 

MR. OLNEY: The question of the frequency range of radio receivers has been 
raised in the discussion several times, and someone inquired as to how it compared 
with the reproduction we heard today. There is no comparison between them. 
In order to reproduce the low frequencies you have been hearing today, a receiver 
cabinet would have to be the size of the panel you see there, which is out of the 
question. As far as high frequencies are concerned, the requirements of selectivity 
prohibit reproduction of anything higher than five thousand cycles. This is a 
theoretical limit. Practically, the response drops off in the best radio receivers 
between four and five thousand cycles. In the poorer receivers it may drop off 



158 H. A. FREDERICK [J. S. M. P. E. 

at three thousand cycles. In radio receivers equipped with the so-called tone 
controls, it may be possible for the user to reduce the cut-off frequency to fifteen 
hundred cycles. Some persons prefer that. 

I do not think it is because they object to a normal amount of high frequencies. 
Some have claimed that what annoyed them were the frequencies above five 
thousand cycles. Those frequencies, I believe, are not reproduced by any radio 
receiving set on the open market today with which adjacent stations can be 
separated. I believe that one of the difficulties is that most loud speakers used 
in radio receivers have a very exaggerated response to frequencies in the neighbor- 
hood of twenty-five hundred or three thousand cycles; and unless these peaks 
are suppressed in some manner, the reproduction is bound to be unpleasant. 

In commercial receivers, correctly designed, an attempt is made to equalize 
these defects in the loud speakers. When they are not equalized you will get 
this impression of harshness in the upper register; but this is not due to the 
frequencies above five thousand. 

MR. VICTOR: I am afraid these discussions will read as if our Society members 
criticized this performance. I should not like to see the records so appear after 
such a splendid performance. It is the best I have ever heard. 

PRESIDENT CRABTREE: Exactly. But we must not close our eyes to the 
imperfections. We will never progress if we do not criticize our own work. The 
man who is satisfied with his own work never gets anywhere. 

MR. FREDERICK: I should like to add a comment suggested by something 
Mr. Victor said, that I may not have properly answered. As the volume level 
of any reproduced sound is raised or lowered, the quality appears to change, 
due to the physiological characteristics of our hearing. If it is played too loud 
the balance is off in one direction, and if it is played softer than in the first place 
the reproduction is off in another direction, even though the frequency character- 
istics of the reproducing system, taken either in part or in whole, are entirely flat. 

Now, one interpretation of his comment might be that if a person simply 
must play the reproduced sounds too loudly, he needs to have a certain amount 
of distortion to make up more or less for what his ear is doing. Of course, 
I do not think that is the proper thing to do. He ought to endeavor to play 
the reproduced sounds at the level at which they were originally produced and 
picked up. If that were done he would not need that kind of a compensating 
system. 

MR. HICKMAN: Mr. Victor put a point for me, that I had in mind. I think 
that anybody in reading this discussion might imagine that our criticism had 
been of this demonstration. This demonstration has been the most perfect 
music I have heard. My own criticism referred to reproduced music in general. 

DR. GOLDSMITH: There were two points about this extremely interesting and 
significant demonstration, which I believe merit consideration: In the first 
place, comparisons were made to radio. There is no possibility of comparison 
to radio, because transmitting stations today send out approximately forty-five 
hundred cycles. The networks of the country carry very little above that. 
A few stations are going up to approximately eight thousand cycles in trans- 
mission. But taking five thousand cycles as the present transmission of radio 
stations, it is obvious that a wide open receiver, receiving from zero to ten thou- 
sand cycles, would get only the upper five thousand cycles of extra noises super- 



Feb., 1932] VERTICAL SOUND RECORDS 159 

imposed on the signal; and a five thousand cycle low pass one would be per- 
fectly justifiable. 

If, below that, there are cut out frequencies from five thousand down to three 
thousand, then reception suffers considerably. A great many people prefer 
that, however, because they have been working in noisy locations all day, and 
want to be soothed rather than given an esthetically true reproduction. 

Then, a corollary of that is that home conditions are not the same as conditions 
in an auditorium. If you listen to a great conductor, you discover that the 
faintest whisper of sound can barely be heard, whereas, the climax following 
immediately, nearly brings the plaster off the ceiling, if the theater has not been 
acoustically treated. That condition must not exist in the home for the reason 
that a radio receiver in the home would be an intolerable nuisance and would 
start a neighborhood feud. We have to limit the volume range. We have 
something here that is adequate, indeed, for home purposes, and has the maximum 
of volume range which is permissible and consistent where people live near each 
other. 

Another feature that we have to take into account is our own reaction when 
we hear things for the first time, in that we have a habit of reverting to old stand- 
ards. I remember the first time a certain man, who was quite a capable man in 
his field, listened to a modern record, and remarked, "That is not good at all. 
It does not sound like a phonograph." And so we have to be careful. We 
must remember that we have all made a mental adjustment, a charitable ad- 
justment if you wish, to reproduced music. We have accustomed ourselves 
to making mental allowances, applying the necessary automatic corrections. 
And we apply them to something where they are no longer relative. 

I regard this as a most impressive demonstration. And throwing a bright 
light on nature, and holding up a brightly polished mirror to look at nature, 
is crude, but it is the only way to progress. 

One other point, and that is the matter of the capabilities of film records. It 
is to be hoped and believed that results like this will properly stimulate the 
production of film records for theater use, which will be of equivalent quality. 
There is nothing theoretical or impossible about that. The trouble now is not 
with what is on the film, but the acoustic qualities of the theater. 

MR. FREDERICK: I think it would be hopeless to try to summarize all the 
discussion which has taken place. 

I regret that I am not fully familiar with all the facts regarding the transmission 
circuits of the networks connecting studios to radio stations. I know a great 
many of them are good to eight thousand cycles, and a great deal of effort has 
been made to make them good to eight thousand cycles. I think it would be 
unfortunate and quite incorrect if we should take away with us the impression 
that the transmission circuits were limiting or will limit radio transmission to 
five thousand cycles. 

(The following discussion was held on the occasion of the re-presentation of Mr. 
Frederick's paper, and the address of Mr. Stokowski published on page 164 of this 
issue of the JOURNAL, at the meeting of the New York Section on December 9, 1931.) 

MR. RICHARDSON: One of the worst things we have to contend with in re- 
production and projection of sound is dust, both in records and in the film itself. 



160 H. A. FREDERICK [J. S. M. p. E. 

It will adhere to the film, particularly when a little oil has gotten on the film 
and will set up heavy ground noise. 

It would seem to me that a record of hills and dales would be much more easily 
injured; it would be more difficult for the projectionist to keep it clear of dust 
and abrasive materials than a lateral record. 

It also seems to me that a sapphire or a diamond point needle running over 
dust, which unquestionably will collect in the grooves, would have a more in- 
jurious effect and set up a greater amount of ground noise than would be the 
case with the lateral record groove. 

Finally, I should like to know what are the limits of range of frequency in con- 
versation and in music. 

MR. FREDERICK: As to the effect of dust, our experience with these records 
has not indicated any particular difficulty. We have taken no particular 
precautions to avoid trouble. Except where we wanted to play a record thou- 
sands and thousands of times, continuously, we have found it necessary to take 
no special precautions whatever. 

As to the limits of speech or music, opinions may differ as to that. If anyone 
will provide an adequate or an accurate audiogram showing what the upper and 
lower frequency limits are for his own ear, that will provide the answer to the 
question. If he is very young and can hear from twenty cycles to seventeen 
thousand cycles, twenty and seventeen thousand are the limits of his speech and 
music. If his hearing isn't quite so good and he can hear only from twenty to 
three thousand, why, that is the limit for him. 

(At Mr. Crabtree's request, Mr. Frederick repeated a record, using only the speakers 
which reproduced the high frequencies.} 

MR. CRABTREE: Mr. Chairman, I hasten to congratulate Mr. Frederick and 
his collaborators at the Bell Laboratories on this epoch-making development. 
The demonstration this evening, especially of the organ, shows how inadequate 
the present apparatus is and the present theory. I don't know whether Mr. 
Zukor, Mr. Lasky, or Mr. Harley Clarke were here tonight but if they weren't 
they should have been, and perhaps it might offer them some hope of getting the 
people into the motion picture theaters today if they would put on music of this 
high quality. 

MR. SCHENCK: In your opinion, Mr. Frederick, when these high frequency 
speakers are playing we haven't been accustomed to hearing frequencies of 
that order were those actual reproductions on that frequency band or would 
you say there was distortion present? 

MR. FREDERICK: There is no question that the day of reproducing sound 
without distortion is not yet here. Surely there is some distortion there. All I 
should say is that I think there is a little less than I have often heard, and I 
hope that five years hence there will be still less. 

I waver between two feelings on this whole matter. Some days I quite enjoy 
listening to some of this music, and most of the other days I feel greatly impressed 
with the fact that we yet have a long, long way to go. This is not perfect and 
the day of perfection is a long way off. 

MR. SCHENCK: We are not accustomed to hearing the high frequencies re- 
produced, and I am merely wondering whether we jump to conclusions about the 
distortion at that high frequency, particularly in connection with the orchestral 



Feb., 1932] VERTICAL SOUND RECORDS 161 

record, wherein it sounded to me as if the cymbals were playing. At one point 
in the record it seemed almost certain that there was distortion. Following 
that, it started to clear up somewhat, and I could hear the high frequency instru- 
ments such as the cymbals or the bells that you mention. 

MR. STOKOWSKI: Those loud crashes are cymbals. But they are cut off at 
nine thousand cycles. You need at least thirteen thousand, according to our 
experiments, and perhaps more. That is why they sounded strange. 

MR. CRABTREE: What is the thickness of the records, and how are they made? 

MR. FREDERICK: These particular records are about a quarter inch thick. 
They can be made two-hundredths of an inch thick. They are thermoplastic, 
not like bakelite. Under the application of heat they soften. They were pressed 
in the usual manner. 

MR. WILSON: I would like to ask Mr. Frederick if the sound level, as it 
appears to the average person sitting here tonight listening to the orchestral 
record, is approximately what would be expected in an equivalent position in an 
auditorium listening to the actual orchestra. It is difficult, looking at a bank of 
loud speakers, to judge whether one is hearing the true level or something con- 
siderably above what he would get from the real orchestra. 

MR. FREDERICK: I fully appreciate the difficulty you express of judging 
whether the level is right. Remember, you always hear an orchestra with two 
ears. The binaural effect changes your impressions always. You also use your 
eyes when you hear an orchestra and I think that what you see also changes the 
general impression quite a lot. 

The loudness can, of course, be definitely measured and can be compared 
under different cases. We haven't actually made such measurements in this 
hall, but it is my impression that the loudness, both in the case of the organ 
record and in the case of the orchestral record, was fairly close to the original 
loudness. 

MR. EDWARDS: I should like to ask Mr. Frederick about tracking. That 
is the great handicap of lateral recording, the thing that has given the most 
trouble in projection. 

This is the first time that I have seen a reproducer that hasn't depended for 
its trackage on a thread and screw. 

In the illustration showing the difference between the hill and dale and the 
lateral recording as placed on the record, I noticed that in the case of the hill 
and dale recording the wall of the record is cut very much lower than the surface 
level. Would not a little wear cause a great deal of difficulty in tracking, es- 
pecially with the free reproducer? 

MR. FREDERICK: I don't think I have ever seen one of these fail to track. We 
have had practically no trouble at all from this. I don't doubt there may be cases 
where they haven't tracked but I don't remember ever seeing one. That hasn't 
been one of our difficulties. 

MR. EDWARDS: I think possibly the most notable example of detracking in a 
lateral record was in the picture, Lilac Time. In that picture there was a shot 
in the center of one record, and I think that shot must have cost the producing 
company a matter of twenty-five thousand dollars for that record because, once 
played, the next time it went through the wall. It brought disastrous results 
to everybody concerned. 



162 H. A. FREDERICK [J. S. M. P. E. 

MR. FREDERICK: Of course, with a lateral record, if an extra broad deviation 
of the groove occurs there is danger of cutting over into the next groove. With 
the vertical cut, even when the cutting stylus leaves the wax entirely, we have 
never experienced any difficulty in tracking. There is some distortion, of course, 
due to the fact that the top of the wave has been cut off. But it tracks perfectly 
well. And I think that is a rather important practical advantage of the vertical 
as opposed to the lateral type of record. 

MR. CRABTREE: Might I ask Mr. Stokowski to tell us what is lacking in the 
music from a musician's standpoint? First of all, is the volume adequate? 
Do you get the thrill from the reproduction that you do from the actual orchestra? 
Is it lacking in depth or static effect? Do you notice the lack of perspective in it? 

It is only by criticism of this kind, of course, that we can really advance; 
find out what is lacking and then try to improve it. 

MR. STOKOWSKI: As to volume range, it is approximately, in my opinion, 
the same as in the original orchestra but in frequency there is a departure. The 
cymbals don't sound like cymbals because, as I said before, they are cut off 
at about nine thousand cycles and they need thirteen or fourteen thousand. 
The range between nine thousand and thirteen or fourteen thousand is necessary 
for several other instruments to give the proper tone color. It is a pity we 
do not have a word in the English language for timbre. We ought to invent 
one, because we need technical terms which have an exact significance and are 
invariable in their meaning. 

We have in Philadelphia, in the monitor room (a room, I suppose, about one 
hundred and twenty feet long, so that there is plenty of space in which the tone 
can develop), the Bell Telephone Laboratories' loud speaker, different from 
this one. This is a double speaker; we have a triple loud speaker there, wired 
from the microphones in the hall. We have usually three, four, or five different 
microphones in different positions, so we can switch from one to the other. 

When we sit in that room, which is soundproof, we don't hear the original. 
We hear the music only from the loud speaker. And we have there a most 
wonderful and faithful reproduction of the orchestra. I go in and conduct the 
orchestra for a time, to get the direct sound of the orchestra. Then I go down 
the hall about two-thirds of the way and listen to the orchestra from that point, 
which is the average listening point for the public. Then I go into the monitor 
room and compare what I hear there with what I heard outside, and it is a very 
faithful reproduction. From that comparison I notice that if we cut off from 
about 15,000, as we have done there, down to 9000, we not only cut off those 
higher frequencies, but there is some relation between those high frequencies 
and the ones which exist from 9000 downward, and they, too, are changed. The 
frequencies ranging from one to five thousand should remain the same when the 
frequency range is cut down to 9000 but, to the ear, they don't remain the same. 
They are changed in some way. You get a totally different sound. And that 
is, I think, one thing that will be gained when we have still higher frequencies 
than this, which we undoubtedly will have, because we already have them in 
Philadelphia. 

MR. RICHARDSON: I believe that this style of recording will meet with trouble 
in the projection room due to the abrasive effect of dust in the bottom of the 
groove where the pressure must come from the needle. It must be borne in 



Feb., 1932] VERTICAL SOUND RECORDS 163 

mind that the conditions in the laboratory and those in a projection room are 
quite different, particularly in the smaller theaters. 

MR. HORNBLOWER: I should like to know whether, in checking the original 
against the reproduced sound of the symphony, consideration was given to the 
fact that the symphony orchestra would occupy a stage as large as the one you 
are standing on, that your base drum would be, say, thirty feet from the first 
viol, and so on, while in reproducing you get everything within an area six feet 
square. 

MR. FREDERICK: I tried to bring out that point before, that one of the limita- 
tions of this type of reproduction is that we are effectively listening with only 
one ear, picking up with a single microphone, whereas under normal conditions, 
in a hall, we hear with both ears, and the orchestra is spread out. 

We have taken that fact into account in some recording work by placing the 
microphone at an adequate distance from the orchestra. 

In the particular orchestral record which we played here, we were obliged to 
have the microphone close to the conductor's stand, which we know is an atrocious 
place for it ; but it was impossible to place it anywhere else, and I am quite sure 
that the record was very much injured as a result of it. 

MR. SIMMIONS: Mr. Chairman, I should like to ask Mr. Frederick why the 
research which has been done at the Bell Telephone Laboratories has been 
confined to the speaking voice and has not included the singing voice. 

MR. FREDERICK: I didn't think it had been. 

MR. SIMMIONS: Then very little has been accomplished in regard to the 
singing voice. 

MR. FREDERICK: Of course, in the telephone business our greatest interest 
is in speech, although we have done some work in other directions. We have 
played records here of singing voices. 

MR. SIMMIONS: I should like to know if the physicist alone can solve the 
problem of the singing voice, from the physical point of view. In order to carry 
on this research work I have suggested that there should be three physiologists, 
three musicians, and three teachers or psychologists of course, the singing teacher 
is a psychologist and then these nine men together could accomplish something 
in regard to finding out the exact amount of pressure which is necessary in order 
to produce a beautiful sound. 

I would suggest Mr. Stokowski, Mr. Damrosch, and Mr. Bodanski as the 
three musicians on that research committee. The Academy of Singing Teachers, 
whom I have approached, suggested that they should select three men from 
among their ranks. Regarding the physiologists, I have spoken to Dr. Williams of 
Columbia University and he is interested. I have spoken to Professor Wisluki, 
who is professor of anatomy at Harvard University, and Dr. Frank E. Miller; 
and to Dr. Fletcher, Dr. Watson, and Dr. Knudsen, of the University of California. 

These men, with the help of the singing teachers, should get together to solve 
the problem. If they did so I am quite sure the problem of singing in relation 
to the films would be solved. But as I say it is not a one-man job. 

The same standards which have been used in checking the singing could be 
applied in teaching control of the human voice, so that I should not have to 
depend on the monitor man when I go on, as Mr. John McCormick does. You 
know, he said the monitor man changes his voice. 



SOUND RECORDING FROM THE MUSICIAN'S POINT OF 

VIEW 

LEOPOLD STOKOWSKI* 



An address delivered before the New York Section, December 9, 1931, following 
the re-presentation of the paper "Vertical Sound Records: Recent Fundamental 
Advances in Recording on Wax," by H. A. Frederick, published in this issue of the 
JOURNAL on page 141. 

As we listen to music, if I may speak purely from the musical 
standpoint, we have two kinds of reaction. First, there is the 
physiological reaction. 

When you heard the great volume of tone coming from the organ it 
thrilled you. It uplifted you. It excited you. I am sure you would 
find upon analysis, that your heart was beating more quickly, that 
your blood was flowing more quickly, that your nervous system was 
tremendously stimulated. That is the physiological reaction. 

If you hear a good military band playing in the street, with a 
really good rhythm, you want to march. That again is physiological. 

If you hear very good dance music, you want to dance. Again, 
the physiological reaction. 

The other kind of reaction is the psychological, the emotional. If 
you hear music of a certain type it arouses in you intense feeling. If 
you hear music which has very powerful contrasts very loud, then 
very soft; very quick, then very slow; and so on that has a 
psychological effect on you. 

If you hear music which has very rich colors in it, and differences of 
tone colors, that again has a psychological effect. Melodic form, the 
flowing up and down of melodies, tunes, motifs that also has a 
psychological effect. 

There is also a very mysterious thing about music. It is psychic 
suggestion. I work all day long and every day in music. I experi- 
ence every day, the whole day, the next day, that week, that month, 
ten years past, this psychic impression and suggestion that comes 
from music. So it is with all musicians; we talk about it; we think 

* Director, Philadelphia Orchestra, Philadelphia, Pa. 
164 



SOUND RECORDING 165 

about it; but we don't know what it really is. We feel it vividly, but 
we don't understand it at all. 

That is, to my mind, the most important part of the reaction of the 
music lover or of any one listening to music. That suggestive power 
which can carry us into the most remote spheres and realms of feeling 
and thought, and things that are higher than thought and higher 
than feeling that is the important part of music. And in order for 
this to be done we must have this greater range which we have had 
demonstrated here tonight; greater range of frequency, of volume, 
and the elimination of foreign noises, needle scratch, static, and all 
the noises that we hear in radio. You hear on your radio the dial 
telephone in the next room; you hear the refrigerator; you can 
hear all the vegetables in the refrigerator talking to each other; and 
when the cook takes them out of the refrigerator and puts them on 
the electric stove and switches it on, you hear that. And so it goes. 
We must find methods of eliminating all foreign sounds. 

When our orchestra plays in Philadelphia, or as we played last 
night at Carnegie Hall, here in New York, we give out a volume range 
of about 75 db. ; but when we are recording we do so at about 35 db. 
And I think it is important for everyone connected with music, and 
the public at large, to know definitely and quite clearly it is no 
secret, but quite plain that when they listen to a record, or when 
they listen to the radio, they are listening to a sound level of approxi- 
mately 35 db., sometimes less, sometimes a little more. But when 
they listen to a symphony orchestra, which is, I think you will agree 
with me, the most difficult thing to record or to transmit, they are 
listening to a level of about 75. We must find a way of increasing 
that 35 to 75 before we really can give the public what it ought to 
have in the way of expression in music. 

That is one dimension, so to speak. Then there is the other 
dimension, the up and down dimension, the frequency range. When 
we play as we did last night at Carnegie Hall, in the overtones, or in 
the fundamentals, we are producing frequencies certainly up to 13,000, 
probably more. But we know certainly that it is up to 13,000. 
When you hear a record or when you hear music over the radio, you 
are hearing frequencies of about 4500, often less, sometimes a little 
more. The average, however, is about that. You can easily measure 
it and find out for yourselves whether I am telling the truth or not. 

Last Friday night we had a concert in Philadelphia, and after the 
concert we made a number of tests, in connection with the Bell 



166 LEOPOLD STOKOWSKI [J. S. M. P. E. 

Telephone Laboratories, and these are the exact figures we got from 
those tests: 

We asked the first oboe player to play. We were in a room a long 
way from the room in which he was playing. We had previously 
arranged everything so that what we heard was an exact reproduction 
of what was happening on the stage. The oboe player was sitting in 
the same seat he always occupies during a concert, so that it was an 
exact reproduction. And we found that he needed frequencies up to 
13,000 to express his tone color. 

Then we took the trumpet and we found that up to 8000 cycles it 
gave a satisfactory effect. 

The piccolo took up to 6000, and that was a very astonishing thing: 
that the piccolo, which is a very high pitched instrument, should 
require up to 6000, whereas the trumpet, which is a lower instru- 
ment, requires up to 8000 and the oboe, a moderately low pitched 
instrument, requires up to 13,000. That is something you couldn't 
determine without exact experiments like these. 

Then we took the violin. It needed up to 8000. The cymbals 
needed up to at least 13,000, probably more; the tympani, 6000; 
triangle, 13,000; xylophone, 6000; snare-drum, 13,000. 

I was doing these experiments, but the Bell Laboratory scientists 
were all watching very closely so that there was no chance of exag- 
geration or mistake. Those are the exact results. 

In order to express all this, in my opinion, we must find out what 
the average living room, with the average curtains, rugs, paintings, 
and all the things that our wives like to have in our living rooms, which 
affect the tone, its absorption, and so forth we must find out what 
the average living room will take in the way of volume range. We 
really don't know that exactly, yet. At least, I have never found 
anybody who did. In my opinion we must know that and we must 
experiment along that line. 

The same thing applies to the average theater in which sound 
pictures are shown. They vary greatly, and when we record sound, 
music, or speech, no matter what kind of sound it is, we must have 
those conditions as nearly as possible invariable. They must be the 
same, because we record in a certain way, to project the sound in a 
certain way, and then if the projecting instrument and the hall or 
room in which it is sounding is different in each case, a different effect 
will be produced in each case. 

This is the place, in my opinion, where standardization is very 



Feb., 1932] SOUND RECORDING 167 

desirable. In many other things in life, such as thought, emotion, 
etc., it is very undesirable to have standardization. But it is impor- 
tant for us to see clearly where standardization is necessary and I 
think it is necessary here. 

It is the same with frequency characteristics. This hall has 
certain frequency characteristics; your living room where you play 
your radio, where you play your gramophone, has frequency character- 
istics. In producing our music for the gramophone or for the radio 
we should know roughly what is going to be the frequency character- 
istic of the place in which it is going to be played or the whole thing 
will be distorted. 

In my opinion the gramophone and radio are twin brothers. There 
is often a certain antagonism between those who follow one god and 
those who follow the other god. But they are fundamentally the 
same, and they help each other very much. 

For instance, we broadcast our concert last Saturday night. 
Forgive me if I speak about what we are doing. I do it with a 
definite purpose, not to be personal or to boast in any way far from 
that but I want to tell you tonight about my own direct experience, 
not what I have read in books or what someone has told me, but what 
I have tried by experiment. I think that is the only thing that has 
any value. 

When we were recording last Saturday night, for example, we sent 
this music out. We asked the public to send us criticisms. That is 
what we want. We want them to tell us what is wrong about the 
broadcast, because we honestly want to make our broadcasting better 
and better all the time. Those criticisms came in, hundreds of letters 
and telegrams, telling exactly what those people felt was not good 
about this thing. 

That is one method we have of checking. Another one is that 
during the performance someone is recording the concert in the 
concert hall where we are playing, and of course the connection 
between the microphone and the recording instrument is close and 
can be well taken care of, so that it is in good order and we have good 
reproduction. 

Then the selection is sent over the air, and in the laboratory in 
New York someone is again recording, over the ether. 

So that we are using those three methods of checking our perform- 
ance and comparing them one with the other. First is the criticism 
from the public what we want is the reaction of the average man in 



168 LEOPOLD STOKOWSKI [J. S. M. P. E. 

the average living room who is listening to our broadcast; we want to 
know how it impresses him, and we are receiving that information 
through the letters. Then we have the recording in the academy, and 
the recording over the ether in New York. By comparing those three 
things we get a fairly clear idea of what is wrong and how perhaps 
we can improve. 

People often say when they listen to music, especially modern 
music, "That isn't music." For example we recently produced an 
opera called Wozzek and one of the music critics wrote this in his 
newspaper (as I say, this was a very modern work, different from other 
works, extremely original): "This department is organized to criticize 
music. Wozzek is not music. Therefore we shall say nothing." 

But what is music? What are the limitations of music? There 
are people who think that the Last Rose of Summer is the summit, the 
highest peak of music. Well, it is a very beautiful melody. I enjoy 
it very much when it is well sung or well played. But there are other 
kinds of music, too. A little bird singing in the forest is producing 
very marvelous music, and a different kind. 

What are the limitations of music in sound? Personally, I think 
one sees, as music progresses and has wider and wider horizons, that 
its limitations are becoming less. We are seeing it in a bigger and 
bigger way all the time. And ultimately it may be that we will think 
that all sound is music. All sound has something to which we can 
respond. 

The sound that comes from that little machine* down there I should 
call music because it has a definite frequency. It has definite dura- 
tion, and it has a very interesting rhythm if you will listen to it. 
The narrow-minded musician would say, "No, that is merely a noise." 

But, I think, for the sake of the motion picture with sound, with tone, 
which is going to be an ever and ever more important type of art, that 
we have to think about what is sound, and what is music, and what 
are the limitations of music; and we have to take in more and more of 
sound, the sounds of nature, like the wind going through the trees 
The sound of the sea has a most interesting rhythm if you will take the 
trouble to listen to it. It has very deep, strange sounds, which are 
quite extraordinary. The sounds of the birds are marvelously 
beautiful as Wagner has shown in Siegfried. 

There are all kinds of sounds in nature which are interesting and 

* The stenotype machine. 



Feb., 1932] SOUND RECORDING 169 

which we wish to reproduce in tone films. The machine has very 
interesting rhythms if you listen to it with an open mind, not with the 
nineteenth century narrow-minded view, but with the view of today, 
which takes in more of life. All this will come, in my opinion, in the 
tone film, and then music will not be a narrow thing. It will extend 
itself until it takes in all sounds. 

I said a little while ago that we must standardize certain things. I 
think a great battle is coming in the world between standardization 
and non-standardization, individuality. It is coming in all planes. 
It certainly is coming in the field of economics. I believe it is coming 
in the things in which we are interested, in sound, in science, in 
photography, in light, and I am watching it with great interest. 

For example, my orchestra, I notice, plays differently every day. 
We play the same music; we rehearse it, say, for five days in succes- 
sion. That is what we do every week. We rehearse every morning 
Monday, Tuesday, Wednesday, Thursday, and Friday the same 
music. Friday afternoon we play it in the concert. But every day 
the men play differently. Each day one can draw from them a 
different quality of tone and a different volume of tone. 

This is going to be the great question: whether we shall standardize 
that or whether we shall allow it to be free and individual. Certainly, 
when we record it we standardize it. We must. We fix that day's 
impression on the disk, and send it out to the receiving apparatus as a 
standardized thing. But when we play it in the concert it is un- 
standardized. It is different every day. Emotionally, it is also 
different. That is something for us to think about, and I believe it 
will be years before we get any results on that. 

I believe that this tremendous development that has been going 
on hi sound in the last six or seven years through the radio, the 
gramophone, etc., will lead to something that is very desirable. 

At the present time, when a composer hears in his inner being 
some music, he desires to make it permanent, that impression that is 
going on within him, so he takes a paper and pencil, and writes down 
marks on the paper to preserve that melody, those harmonies, those 
rhythms. Then the singer or the player comes and reproduces those 
sounds. The composer listens and he says, "That is not at all what I 
intended." 

We have that all the time. He composes something for the 
orchestra. We play the notes that he has written and he says, 
"That is totally different from what I intended." 



170 LEOPOLD STOKOWSKI [j. s. M. P. E. 

Why is that ? It is because the method of writing sounds on paper 
is tremendously imperfect. 

If a painter wishes to paint a picture he takes his canvas and his 
colors; he puts his colors on the canvas where he wishes them. He 
makes his design, his relativity of color to color or form to form, 
and when he finishes it and he is satisfied; that ends the matter. 
It is complete. 

But when Beethoven or any composer composes a symphony and 
writes it on paper, he has only half completed the process. It must 
then be given to the orchestra. They play it and he is dissatisfied, 
because it doesn't reproduce his idea, because our method of notation 
is so imperfect 

I see in all this development something new coming. I believe it 
will be only a few years before the composer will paint directly in 
tone. He won't write down his impressions on paper. He will 
express them through frequency, through volume, and through 
duration. In that way he will express his ideas exactly, and not with 
the imperfections we now have. That is almost possible today, and 
through electrical production of tone, such as we get through the 
Theremin instruments and others which are being developed now, 
that will soon be possible and will be a very desirable thing. 

What is the ideal for us who are scientists, or engineers, or musicians 
or photographers, or producers of tone films? What can we do in 
the future which is greater than what we are doing now? A great 
deal, in my opinion. 

We may communicate with someone by telephone. We can talk to 
someone over the telephone. We can communicate ideas. We can 
come to understandings about ideas. We can talk for a long time on 
the most intricate, complex subjects, and make decisions and have a 
discussion. But when we combine sight and sound, through the 
tone picture, we can communicate much more, not only ideas but 
emotions and suggestions of things which are not completely said but 
which are conveyed in a more subtle way. We can suggest on levels of 
consciousness higher than thought, and feeling, and imagination, and 
all those strange things that go on in our nervous system which make 
our inner life so complex and so rich. Above all, these things for which 
we have words we all know perfectly well there are other things. We 
have no names for them, no words for them, but they exist. They 
are part of our daily experience. Especially do we feel those things 
through the finest type of music. Music of the higher type expresses 



Feb., 1932] SOUND RECORDING 171 

just those things. And it is through the tone film that we can very 
richly and completely express that, and it is through radio, and 
eventually television that we can project those things through space 
all over the world. 

That is the magnificent ideal, something quite supreme, toward 
which we must all work. We must not be satisfied to stand where 
we are at present, which is about a half-way point toward that thing. 

The development of the radio, the gramophone, of photography 
and reproduction of sound has been perfectly miraculous during the 
last six, eight, or ten years, but there is far more yet to be done. 
Let's admit that frankly, and let's work for that immense ideal which 
is possible. 



ON THE ASSIGNMENT OF PRINTING EXPOSURE BY 
MEASUREMENT OF NEGATIVE CHARACTERISTICS* 

CLIFTON TUTTLE** 



Summary. The theory of photographic tone reproduction, though specific for 
ideal cases, cannot always be applied in the determination of printing exposure for 
motion picture negatives. A statistical study of the correlation of various optical 
characteristics maximum transmission, minimum transmission, and total frame 
transmission with the required exposure has been made. Of the possible measure- 
ments to be made, the value of total frame transmission seems to be the best criterion 
of printing exposure. The apparatus used in making the measurements is described 
and the data obtained are presented graphically. 

PRESENT PRACTICE IN PRINTING EXPOSURE ASSIGNMENT 

In motion picture finishing laboratories, one of the problems 
which must be considered is the assignment of printing exposure 
to each scene of negative of which a print is desired. The usual 
type of printer operates at constant speed, thus fixing the time of 
exposure. Compensation for differences in negative density is 
made by varying the intensity of the light incident upon the negative. 
In practice, a series of intensity steps is provided either by control 
of resistance in series with the printing lamp or by the setting of an 
opening in an optical diaphragm. Before a negative is printed its 
correct printing intensity must be selected and the light source 
must be regulated to give this intensity. 

Methods for selecting the best printing exposure vary somewhat 
in different laboratories. In some instances, a tablet sensitometer 
is used as described by Jones and Crabtree. 1 In this method, a 
print of the negative scene is made through a density step tablet. 
The steps, each the size of a single frame, have been calibrated to 
correspond with those of the printer light-change board. The 
resultant positive after processing is inspected visually and a se- 

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. Communica- 
tion No. 470 from the Kodak Research Laboratories. 

** Research Laboratories, Eastman Kodak Co., Rochester, N. Y. 
172 



ASSIGNMENT OF PRINTING EXPOSURE 173 

lection of the best exposure is made by an expert judge of print 
quality. 

More frequently the assignment of printing exposure is made 
directly from the negative. A negative "timer," who by virtue of 
long experience and particular aptitude has become adept at judging 
negatives, is able to assign the proper printing exposure to a series 
of negative scenes merely by visual inspection. In the timing of 
negatives used for release prints, the initial results obtained by the 
timer are, of course, subject to correction after the projection of a 
trial print. 

The author has never had the opportunity to gather any data 
concerning the waste of time and material occasioned by errors 
in the initial timing of a negative. It is probable, however, that 
such waste amounts to a negligible per cent of the total processing 
cost on release pictures, and it is probably true that the present 
methods of exposure assignment are entirely satisfactory where a 
large number of prints are made from a single negative. 

If a single print is to be made from a negative, if speed is desired 
in the production of a first print, or if the services of an expert judge 
of photographic quality are not available, assignment of printing 
exposure on the basis of a measurement of the optical character- 
istics of the negative may be desirable. These practical considera- 
tions and the obvious interest of the question in the theory of tone 
reproduction have suggested the value of a study of the relation of 
the optical characteristics of motion picture negatives and their 
required printing exposures. 

NEGATIVE CHARACTERISTICS AND PRINTING EXPOSURE 

The rigid theory of tone reproduction is specific on the subject 
of required printing exposure. To reproduce with perfect accuracy 
the brightness relationships existing in the object by an equivalent 
series of tone relationships in the picture requires, first, a negative 
in which the total range is included on the straight line portion of the 
H & D characteristic curve. Given such a negative, a perfect 
print must translate the negative density range into positive density 
values lying in inverse order on the straight line portion of the 
positive characteristic curve. For the thinnest perfect positive, 
which for efficient projection would seem to be the thing desired, 
the printing intensity, according to the tone reproduction theory, 
is given by the following: 



174 CLIFTON TUTTLE [j. s. M. P. E. 

lOg / = log E P mia. - log t -h D N max . 

where / = intensity incident upon the negative. 

/ = time of exposure. 
DN nu. = maximum negative density. 

EP mm. = minimum exposure for positive the exposure for the lowest 
density on the straight line portion of the characteristic. 

In practice, many negatives, probably most of them, are not 
perfect in the sense just described. The printing operation also 
is usually a compromise, and throughout the literature we find 
numerous suggestions as to practical criteria for printing exposure. 

Hurter and Driffield, the pioneers of quantitative photography, 
make the following statement in one of their early papers: 2 "We 
first of all measure the highest density of the negative . . . and 
knowing the inertia of the plate (positive) we take care that the 
exposure shall be such that behind the highest density of the negative 
the plate shall receive an exposure at least equal to the inertia." 
Since the inertia is defined as the intersection of the extended straight 
line portion of the characteristic curve with the log R axis, it is evi- 
dent that an exposure equal to the inertia will not give a density 
lying on the straight line portion of the positive characteristic. 

Driffield, in a later paper, 3 modified this criterion. He suggested 
that the printing exposure be computed as the antilog of the average 
of maximum and minimum negative densities multiplied by the 
geometrical mean of the exposure range of the positive. This pro- 
cedure bases the printing exposure on the transmission of the middle 
tones of the negative. 

According to this criterion, a negative with a range of density 
greater than can be accommodated by the linear portion of the 
positive characteristic would give a print in which shadow and high- 
light would overlap shoulder and toe of the positive characteristic 
to the same extent. Since that time a number of others have recom- 
mended the use of a similar criterion of printing exposure. 

One authority quoted by Renwick 4 considers the total range of 
the print and not the relationship between tones to be the important 
thing, which is equivalent to saying that printing exposure is in no 
way critical so long as a given maximum contrast is obtained. 

F. C. Tilney, 5 speaking for the artist, remarks, "There seems to 
be one thing only in matters of tone that is absolute, and that is 
the correct relation of one tone to another in the same picture what- 
ever the key adopted." Translated into photographic parlance 
this statement may be taken to mean that the straight line portion 



Feb., 1932] 



ASSIGNMENT OF PRINTING EXPOSURE 



175 



of the characteristic curves only should be used and that the locating 
of negative density values with respect to the log E axis of the 
positive is unimportant so long as this condition is fulfilled. 

If we may be permitted to apply the practical photographer's 
axiom for negative making "Expose for the shadows and let the 
highlights take care of themselves" to the making of the positive, 
we should do well to base our judgment of exposure on the trans- 
mission of the thinnest (or shadow) portion of the negative. 

L. A. Jones' 6 discussion of tone reproduction, which is based upon 
a knowledge of the limitations of photographic materials, fixes the 
printing exposure at the value which will give a "just perceptible" 
density (0.008) for the highlight portion of the negative. This 
procedure insures gradation throughout the highlights of the picture. 




FIG. 1. A projection densitometer for the measurement of motion picture 
image characteristics. 

The foregoing comments indicate that there are some differences 
of opinion regarding the assignment of "best printing exposure" 
from a consideration of negative characteristics. It is hoped that 
this fact will supply an excuse for the statistical treatment of a 
problem which does not appear amenable to purely theoretical 
solution. 

MEASUREMENT OF OPTICAL CHARACTERISTICS OF A NEGATIVE 

Apparatus. Three characteristics of a negative minimum trans- 
mission, maximum transmission, and total transmission are readily 
measurable. Any one of these, or a combination of two, might be 
expected to give some correlation with the required printing exposure. 
To facilitate the measurement of these three values for a large 
number of motion picture negatives the instrument shown dia- 
grammatically in Fig. 1 was constructed. 



176 CLIFTON TUTTLE [j. s. M. P. E. 

Referring to this figure, a monoplane filament lamp, A, ma. suitable 
housing is imaged by lens, B, in the plane, C. In the plane, C, a 
sliding carrier containing a lens, D, and thermopile, E, may be so 
positioned that either the lens or the thermopile receives the filament 
image. 

Over the condenser lens, B, is placed a rectangular mask with an 
opening the size of a single motion picture frame. The aperture 
is supplied with a spring gate so that a motion picture film may be 
readily inserted and framed in the opening. 

When the lens, D, is in position an image of the motion picture 
frame may be formed either at F or at F', depending upon the position 
of the totally reflecting prism, G. This prism may be rotated about 
a vertical axis through its hypotenuse face to either of the two 
positions shown in Fig. 1. The plane at F' is provided either with 
a ground glass screen for viewing the image or with a sheet of bromide 
paper for making a permanent record of it. A stylus back of the 
plane F' may move either in the vertical or horizontal direction so 
that it is possible to indicate any area of the image. The movement 
of the stylus in the plane F' is mechanically linked to the movement 
of the Moll thermopile in the plane F. Thus, if the stylus is posi- 
tioned at an area of the image corresponding, say, to the most dense 
portion of the negative highlight, the thermopile is brought auto- 
matically to an identical position with respect to the image which is 
projected on F by the rotation of the prism. 

The thermopile is connected to a Leeds and Northrup high sensi- 
tivity galvanometer (17 mm./MV.) which is provided with an Ayrton 
shunt. Two readings are required in the making of a transmission 
measurement: A value for zero density, and a second value of the 
amount of light which has passed through the area of the negative 
to be measured. For the first value, the motion picture frame is 
removed from the beam by sliding the whole aperture plate and gate 
assembly horizontally in a pair of gibs. It is possible with this 
instrument to read transmission values as low as 0.1 per cent with 
an error less than 5 per cent, while higher transmission can be read 
to a much higher degree of accuracy. 

Procedure Followed to Obtain Data. Through the courtesy of a 
number of studios on both the west and east coasts, about 1000 
clippings from release picture negatives were obtained. A wide 
variation in subject-matter and composition was represented by 
these samples. From each of the negatives a sensitometric tablet 1 



Feb., 1932] 



ASSIGNMENT OF PRINTING EXPOSURE 



177 



print was made to be used subsequently in the determination of 
required printing exposure. 

A single frame of each scene sample was registered in the gate of 
the motion picture densitometer. With the thermopile, E, centered 
with respect to the frame, a measurement was made of the per- 
centage of light transmitted by the whole frame of the negative. 
It should be noted at this point that the measured value of trans- 



25 



o 



&'5h 

(S> 

LU 
Z 



o 
d 10 - 



fes^ 

K 
Z 



10 80 30 40 50 60 70 6O 90 

DIFFUSE TRANSMISSION OF THE WHOLE FRAME IN PER CENT 

FIG. 2. Statistical summary of distribution of inspected studio negatives 
according to the whole frame transmission. The areas represent the com- 
parative numbers of negatives to be found within each region of transmission. 

mission so obtained is a specular value, and therefore is not identical 
with the diffuse transmission value. This matter must be con- 
sidered in an application of the data to the contact printing problem. 
In order to make measurements of the transmission of the densest 
and thinnest portions of each frame, the lens, A was used to project 
an image of the negative magnified 10 times. With the prism, G, 
positioned to throw this image on the ground glass at F' t the areas 



178 CLIFTON TUTTLE [J. s. M. P. E. 

selected for measurement were designated by the indicator stylus, 
the movement of which automatically positioned the thermopile 
to receive the identical area when the prism was rotated through 
90 degrees. The blackened receiver of the thermopile covered a 
circular area 1.0 centimeter in diameter which corresponded to a 
circle on the negative film of 1.0 millimeter diameter. 

In the measurement of the total transmission of the large number 
of negatives, it soon became apparent that the great majority of 
professionally photographed and processed scenes occupied a rela- 
tively small portion of the possible transmission range for printable 
negatives. The data presented graphically in Fig. 2 is of interest 
in that it indicates the remarkable uniformity of the product of a 
number of studios so far as average density is concerned. 

In the plotting of Fig. 2, the specular transmissions obtained 
directly from the galvanometer readings have been transformed to 
diffuse transmissions* to make the results more, directly applicable 
to the contact printing problem. The figure shows the distribution 
of the per cent of the total number of scenes measured among various 
regions of transmission. It is seen from this figure, for instance, 
that 50 per cent of professional negatives have a total transmission 
of from 20 to 30 per cent and that about 95 per cent have a trans- 
mission between 10 and 60 per cent a range of but 6 to 1. 

An expert judge of print quality working from the sensitometric 
tablet prints assigned the printing exposure for the 1000 negatives. 
An analysis of his results showed that the required printing intensity 
range for 95 per cent of the negatives also covered a range of only 
6 to 1. 

Because of this great preponderance of the available samples 
within these narrow limits of printing exposure and transmission, 
it was decided to select a limited number of negatives distributed 

* The author has shown in a previous paper 7 that the relation between specular 
and diffuse density is of the form D \\ (specular density) = KD-\\- M (diffuse 
density). For motion picture negative film K 1.37 and M = 1.088, approxi- 
mately. In computing the value of diffuse density of picture negatives from 
the specular density in this manner there are three possible sources of error: 
(1) The constants given apply to a truly specular optical system; (2) the values 
of the constants vary somewhat for different emulsions; (3) in substituting a 
value of D \\ in the exponential relation one must assume that the density is 
uniform which is, of course, not the case with a motion picture image. It is 
believed that none of these errors is of any great importance for the type of data 
to be presented. 



Feb., 1932] 



ASSIGNMENT OF PRINTING EXPOSURE 



179 



more uniformly throughout a greater printing range rather than to 
encumber the graphical presentation with data for the entire group. 
The table summarizes the data for the selected series. 

TABLE I 

Table Showing Characteristics of Group of Motion Picture Negatives 



Scene 
Number 


Total 
Specular 
Trans- 
mission 


Total 
Diffuse 
Trans- 
mission 


Maximum 
Specular 
Trans- 
mission 


Maximum 
Diffuse 
Trans- 
mission 


Minimum 
Specular 
Trans- 
mission 


Minimum 
Diffuse 
Trans- 
mission 


Ratio 
Maximum 
to 
Minimum 
Trans- 
mission 


Printing 
Exposure 
M.C.S. 


1 


57.0 


69.0 


69.0 


80.0 


17.0 


28.0 


2.9 


1.03 


2 


56.0 


68.0 


69.0 


80.0 


12.0 


22.0 


3.6 


1.03 


3 


41.0 


55.0 


62.0 


74.0 


7.8 


15.0 


4.9 


1.67 


4 


34.0 


47.0 


62.0 


74.0 


5.9 


12.0 


6.2 


1.67 


5 


30.0 


44.0 


48.0 


62.0 


6.9 


14.0 


4.5 


2.28 


6 


29.0 


42.0 


57.0 


71.0 


6.2 


12.0 


5.9 


1.67 


7 


25.0 


38.0 


49.0 


62.0 


3.8 


8.3 


7.5 


2.28 


8 


20.0 


32.0 


46.0 


60.0 


15.0 


25.0 


2.4 


1.03 


9 


17.0 


28.0 


40.0 


54.0 


5.1 


11.0 


4.9 


2.65 


10 


12.0 


21.0 


26.0 


39.0 


4.8 


10.0 


3.9 


3.64 


11 


11.0 


20.0 


18.0 


29.0 


4.8 


10.0 


2.9 


2.65 


12 


11.0 


20.0 


19.0 


31.0 


1.0 


2.8 


11.0 


3.64 


13 


8.0 


15.0 


17.0 


28.0 


1.1 


3.1 


9.0 


5.05 


14 


7.8 


15.0 


26.0 


39.0 


0.9 


2.5 


15.6 


3.64 


15 


6.9 


14.0 


13.0 


22.0 


1.6 


4.2 


5.2 


5.05 


16 


6.5 


13.0 


11.0 


20.0 


0.8 


2.7 


7.4 


5.05 


17 


6.5 


13.0 


15.0 


25.0 


0.5 


1.6 


15.6 


5.05 


18 


6.5 


13.0 


12.0 


22.0 


0.3 


1.1 


20.0 


5.05 


19 


4.6 


9.8 


. . . 


. . . 


. . . 




. . . 


9.55 


20 


3.2 


7.2 


5.5 


11.0 


0.2 


0.9 


12.2 


13.60 


21 


3.0 


6.7 


2.4 


5.7 


0.4 


1.2 


4.7 


7.70 


22 


2.5 


6.0 


2.7 


6.5 


0.17 


0.8 


8.1 


13.60 


23 


2.4 


5.9 


2.1 


5.1 


0.4 


1.2 


4.2 


13.60 


24 


2.4 


5.8 


3.7 


7.8 


0.3 


1.1 


7.1 


13.60 


25 
26 


1.8 
1.6 


4.6 
4.1 


1.6 


4.2 


0.16 


0.8 


5.2 


25.00 
13.60 


27 


1.5 


4.0 


1.9 


4.8 


0.4 


1.2 


4.0 


31.60 


28 


1.5 


4.0 


1.1 


3.1 


0.07 


0.3 


10.0 


25.00 


29 


1.3 


3.4 


0.7 


2.0 


0.1 


0.4 


8.0 


19.00 



This table is probably self-explanatory with the following brief 
enumeration of the methods of obtaining each column of figures. 
Column 2 gives the ratios of galvanometer deflections with and 
without each negative scene in place. In this case a lens in the 
plane of the frame forms an image of the densitometer lamp on the 
thermopile and the reading is, therefore, a specular measure of the 
transmission of each whole frame. In column 3, the values of 



180 



CLIFTON TUTTLE 



[J. S. M. p. E. 



specular transmissions have been converted to diffuse transmissions 
as explained in the preceding footnote. Column 4 gives the specular 
transmission of the least dense area in each scene. Column 5 shows 
these same values converted to diffuse transmissions. Columns 
6 and 7 give similar data for the specular and diffuse transmissions 
of the densest negative areas. Column 8 lists the ratios of the values 



1.5 



1.2 



'-1.7 




.3 .6 .9 \.Z 

LOG PRINTING EXPOSURE (M.C.5.) 



1,5 



FIG. 3. Relation between login of required printing 
exposure and logio of minimum negative transmission, 
diffusely measured. 

of column 5 to those of column 7. These ratio values are of interest 
in a consideration of the exposure scale of the positive material 
which is to be used. The final column is the result of the expert's 
judgment concerning the printer step required to print each negative 
scene, the positive being developed to a gamma of about 1.6. These 
data are given as the exposure in meter candle seconds which would 



Feb., 1932] 



ASSIGNMENT OF PRINTING EXPOSURE 



181 



be required to print each scene. A calibration of the printer to 
which the "required step" data applied was made by methods of 
photographic photometry, the procedure for which has been pre- 
viously described by the author. 8 The intensity factor of the 
exposure was measured by its photographic effect on positive film 
compared to the effect of a source operated at 5000 degrees K. 




.o .3 .6 .9 \.Z 

LOG PRINTING EXPOSURE (M,C. 5.) 

FIG. 4. Relation between login of required printing ex- 
posure and the mean of the logsu, of maximum and minimum 
negative transmissions. 

In Figs. 3, 4, 5, and 6, logic of required printing exposure as se- 
lected by the expert is shown as the abscissa axes. Figs. 4, 5, and 
6 test the various printing exposure criteria which have been enum- 
erated in the early part of this paper. In Fig. 3, logio of minimum 
negative transmission is plotted. This tests the exposure criterion 
suggested by Hurter and Driffield 2 and by Jones. 6 If some definite 



182 



CLIFTON TUTTLE 



[J. S. M. P. E. 



highlight density is to be produced in the print to make the best 
positive we must assign any departure of the points from a straight 
line of unit slope to the uncertainty of the judgment of the expert. 

The criterion suggested by Driffield 3 is tested in Fig. 4. If we 
suppose that an average of highlight and shadow negative densities 
is to be rendered by a definite positive density this data should define 
a straight line of unit slope. 



a 1,5 
g 



.9 



.6 



\ 



\ 




\ 



X 



\ 



.3 ,<o .9 1-2 1-5 

LOG PRINTING EXPOSURE (M.C.5.) 

FIG. 5. Relation between logic of required printing ex- 
posure and logio of maximum negative transmission. 

Fig. 5 shows logio of maximum transmission, suggested by the 
photographer's rule of exposure and Fig. 6 indicates the correlation 
to be expected from a measurement of total frame transmission. 

A value of total transmission can, of course, be determined much 
more readily than can the value either of minimum or maximum 
transmission. From the point of view of convenience and speed 



Feb., 1932] 



ASSIGNMENT OF PRINTING EXPOSURE 



183 



it would be the best measurement to make for the assignment of 
printing exposure. It seems possible that some total transmission 
measurement other than for the whole frame might give even better 
results and still have the advantages of speed in making the deter- 
mination. It is possible, for instance, that the accuracy might be 
increased if only the foreground of the picture were measured, thus 



1.6 



1.5 






S-: 

L. 

u_ 

Q 






,3 .> .9 , 1.2 

LOG PRINTING EXPOSURE (M.C.5.) 



1.5 



FIG. 6. Relation between logio of required printing ex- 
posure, and logio of total negative transmission. 

leaving the area usually occupied by sky in exteriors and ceiling 
in interiors out of consideration. This alternative was tried but the 
results obtained were disappointing. The correlation of this measure- 
ment with required printing exposure was not nearly as high as was 
the whole frame transmission value. 

A second alternative was tried using a circular mask of 0.75 
inch diameter centrally located with respect to the film frame. 



184 



CLIFTON TUTTLE 



[J. S. M. p. E. 



This idea was followed out on the supposition that the center of 
interest, and therefore the area most desirable to measure, usually 
occupies an area toward the frame center. In the case of these 
measurements, the correlation was somewhat better than that shown 
by any of the other measurements. It should be pointed out, 



u 
or 

<D 



CD 

|. 

a 
a 



o 




,6 



.9 



1.2 



LOG SELECTED PRINTING EXPOSURE 

FIG. 7. Relation between logsio of printing exposure as selected by two expert 
judges of print quality. 

however, that, while statistically the central area measurement may 
give the best correlation, the occasional errors due to grouping of 
the subject interest at the edges of the frame may be of greater 
magnitude than would ever occur in case the whole frame were 
measured. 



Feb., 1932] ASSIGNMENT OF PRINTING- EXPOSURE 185 

Certainty of Exposure Assignment by Visual Judgment. In con- 
sidering the data of Figs. 3 to 6, it should not be assumed that the 
value of required printing exposure chosen by the expert for each 
negative scene is the "correct" value. Undoubtedly, at least in 
the case of some scenes, this value may vary somewhat and still 
result in passable prints. It has already been suggested that the 
judgment of printing exposure is regarded as somewhat of an art by 
the motion picture profession. If this is the case, it may be that 
such factors as personal taste of the observer, and conditions of the 
observation play some part in the selection. 

While it is difficult to arrive at any decision on this question of 
how accurate is the work of the negative timer, the following set 
of data may throw some light on this matter. In our own processing 
laboratory two individuals have had considerable experience in 
assigning printer exposure from inspection of sensitometric tablet 
prints. These two persons work interchangeably and it is generally 
agreed that both are expert judges of print quality. 

After the first timer had completed his work with the tablet prints 
the second was given the same set of prints and asked to assign 
the printing exposures. The diversity of opinion which is indicated 
in Fig. 7 is rather surprising. The two axes of the graph are used 
for the log of printing exposure assigned to the series of negatives 
by the two experts. If the agreement in all cases had been perfect 
all of the points would lie on a straight line of unit slope. 

The figure represents data for 180 scenes. The numbers in the 
circles show the number of scenes which determine the location of 
each point. For only 64 scenes is the agreement of the two observers 
perfect. The remainder of the observations is distributed through- 
out an area which is enveloped by the two dotted lines. To include 
all the scattered points these lines are drawn at positions 0.3 in 
logio E removed from the mean straight line. This means that, at 
least in the case of some of the scenes estimated, there is a printing 
exposure tolerance equal to a factor of two or one-half. The sta- 
tistical method does not reveal the presence of some scenes in which 
conceivably there is very little tolerance in printing exposure. 

DISCUSSION OF RESULTS 

Accuracy Demanded in the Selection of Printing Exposure. The 
wide tolerance in the choice of printing exposure which is suggested 
by the data of Fig. 7 is surprising in view of the known facts con- 



186 CLIFTON TUTTLE [j. s. M. P. E. 

cerning existing practice in the commercial laboratories. Consider 
for a moment the usual light change scale of the production labora- 
tory printer. Few of them are calibrated to accommodate an 
intensity range of more than ten or twelve to one. This range is 
split up into twenty-odd steps and the average magnitude of a 
step is between 10 and 15 per cent. A printing exposure tolerance, 
such as that indicated in Fig. 7, would correspond to plus or minus 
perhaps half a dozen such steps. The experts in the laboratories 
presumably work to a tolerance of plus or minus one printer step. 
The author is in no position to express an opinion concerning the 
desirable accuracy of printing exposure assignment but merely 
wishes to present the following facts which may have some bearing 
on the question. 

In column 8 of the table are given the transmission ratios, maxi- 
mum to minimum, of the studio negatives which were examined. 
These ratios vary from 2.9 to 20.0. It is probable that 20 is an 
extreme case. Special precautions 9 must be observed to obtain 
a lens image brightness ratio for highlight to shadow of more than 
25 to 1.0. With negative developed to a gamma of 0.5 or 0.6, the 
transmission ratio will seldom exceed 15.0. Since the average 
positive material at a gamma of 2.0 has an exposure scale of ap- 
proximately 60 to 1.0 there would appear to be a latitude in printing 
exposure of two or one-half from the mean value without making 
use of the toe or shoulder of the positive characteristic. In other 
words, positives which would render the negative tones perfectly 
could be made from most negatives throughout a four to one range 
of printing intensity. Such positives would differ from each other 
only in average transmission. 

The amount of light reflected to the audience from the screen is 
known to differ in various theaters. The public at present sees 
motion pictures under so many different conditions in different 
theaters, that it seems quite possible that within wide limits the 
average transmission of the positive is a matter of small consequence. 

Choosing the Negative Characteristic to Measure. Whether or not 
measured values can be as satisfactory as expert judgment in as- 
signing printing exposure, it is conceivable that there may be appli- 
cations in the processing laboratory for a quick approximation of 
exposure such as would be afforded by a densitometric method. 
With this end in view we can consider the relative merits of the 
criteria tested in Figs. 3 to 6. 



Feb., 1932] ASSIGNMENT OF PRINTING EXPOSURE 187 

The relations between printing exposure, E, in meter candle 
seconds and negative transmissions in per cent obtained from Figs. 
3, 4, 5, and 6 follow: 
12.9 

(1) E = 7,0.79 m which T min is the diffuse transmission of the 

* min. 

negative highlights. 
33 

(2) E = ,w).89 in which r average is the geometrical average 

* average 

of the diffuse highlight and shadow transmission. 
107 

(3) E in which jT max< is the diffuse transmission of the 

* max. 

negative shadow. 
67 

(4) E = ~ in which r total is the diffusely measured total 

* total 

transmission. 

The measurement of minimum negative transmission has little 
to recommend it. The possible error which would follow its use 
(roughly indicated by the distance separating the dotted lines) is 
greater than that for the other suggested values. The dotted lines 
which designate the area required to include all points are 0.8 in 
log E apart. This separation corresponds to an exposure factor 
of 6.3 which means that a departure of 3.1 times hi exposure from 
the value picked by the expert might be made. The relation in- 
volved is exponential, which means that a linear calibration curve 
between opacity (1/7") and required exposure could not be used. 
In addition to these objections, there is the fact that the lower the 
transmission value the more difficult is the measurement to make 
and the greater is the probability for error in the measurement. 

The other suggested criteria appear to be almost equal in that 
the maximum departure from the visually selected printing exposure 
would be by a factor of about two or one-half. The geometrical 
mean of maximum and minimum transmission, which gives slightly 
better correlation than the other criteria give, is probably ruled out 
as a practical measure of required printing exposure because two 
selected areas would have to be measured and a computation made 
before this value could be applied. 

There is no question but that the value of total frame transmission 
is the most readily applicable to the speedy determination of printing 
intensity. In many instances, no doubt, exposure assignment on 
the basis of a total transmission measurement would be considerably 



188 CLIFTON TUTTLE 

in error. In any number of conceivable cases where the object of 
principal interest occupies a relatively small portion of the frame 
against a background of a markedly different transmission, the total 
transmission will not give an indication of the best printing exposure. 
In scenes where special effects are to be obtained by over- or under- 
printing no generally applicable method of printing intensity evalua- 
tion by measurement is conceivable. 

The value of a measuring method used either alone or to supple- 
ment the judgment of an expert is a matter which must be decided 
upon evidence gathered under practical conditions of operation. 

It is certain that there is waste of some time and material with 
the present methods of printing exposure assignment in the making 
of first prints. Only extensive trials of the possibilities of exposure 
assignment by measurement can decide as to its relative merits. 

REFERENCES 

1 JONES, L. A., AND CRABTREE, J. I.: "A New Sensitometer for the Deter- 
mination of Exposure in Positive Printing," Trans. Soc. Mot. Pict. Eng. (1922), 
No. 15, p. 89. 

2 HURTER, F. H., AND DRiFFiELD, V. C. : "Relation between Negatives and 
Their Positives," J. Soc. Chem. Ind., 10 (Feb. 28, 1891), p. 98. 

3 DRIFFIELD, V. C.: "The Principles Involved in the Calculation of Ex- 
posures for Contact Prints on Bromide Paper," Brit. J. Phot., 40 (1893), p. 606. 

4 RENWICK, F. F. : "Tone Reproduction and Its Limitations," Phot. J., 56 
(n. s. 40) (1916), p. 222. 

5 TILNEY, F. C.: "The Appeal of the Picture," p. 49. 

6 JONES, L. A. : "On the Theory of Tone Reproduction with a Graphic Method 
for the Solution of Problems," /. Franklin Inst., 190 (1920), p. 39. 

7 TUTTLE, CLIFTON: "The Relation between Diffuse and Specular Density," 
J. Opt. Soc. Amer., 12 (1926), p. 559. 

8 TUTTLE, CLIFTON: "Illumination in Motion Picture Printing," Trans. Soc. 
Mot. Pict. Eng., 12 (1928), No. 36, p. 1040. 

9 TUTTLE, CLIFTON, AND WHITE, H. E.: "Factors Which Affect the Contrast 
of a Lens Image in the Motion Picture Camera," Trans. Soc. Mot. Pict. Eng., 
11 (1927), No. 31, p. 591. 



UTILIZATION OF DESIRABLE SEATING AREAS IN RE- 
LATION TO SCREEN SHAPES AND SIZES AND 
THEATER FLOOR INCLINATIONS * 

BEN SCHLANGER ** 



Summary. The aim of this paper is to establish a relation between the bodily 
posture of the viewer, the size and shape of the picture, and the architectural form of 
the theater in all its details. The present type of theater floor is compared with the 
reversed type described in a previous paper in order to show how the latter type of floor 
permits placing a greater number of seats within the desirable seating areas than the 
present type. An analysis is made also of the effect of reversing the floor on the 
ability of the viewer to assume a comfortable bodily posture. Definite angles of sight 
specified by the various tilts of chair backs found necessary for comfortable posture 
are shown. Several forms of theaters of various seating capacities and screen sizes 
are described in order to show the broad application of the theories involved in reversing 
the pitch of the orchestra floor. 

The principle of reversing the slope of the orchestra floor in theater 
structures, as presented in a previous paper, suggested the possibility 
of correcting many of the faults of present-day theaters. Bodily 
posture in seating, vision, projection angles, accessibility of various 
levels, and construction costs are all affected. Further study of this 
new principle in planning theaters has resulted in the development of 
definite relations between the various functions that contribute to the 
practicability of the whole. Study has also brought out the fact that 
this new principle is not only applicable for improving the present 
form of the theater, but also for deriving from it many new forms more 
adaptable to motion picture exhibition. (Fig. 1.) 

A complete analysis of bodily posture has been made in connection 
with this new principle. Certain maximum and minimum pitches 
of chair backs and floor slopes have been arrived at, and measure- 
ments have been made of the vertical range of vision which can be 
obtained while sitting against differently pitched chair backs. 

Practical projectionists have verified the need of lessening the angle 
of projection. This need has been recognized, and has been answered 

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. 
** Architect, New York, N. Y. 

189 



190 



BEN SCHLANGER 



[J. S. M. P E. 



in these studies by establishing a maximum angle of ten degrees to 
the center of the screen from the lens center. In most cases the angle 
will be less, varying from ten degrees to a perfectly horizontal line 
of projection. In existing theaters the projection angle is often as 
great as thirty degrees or more. Regardless of the size or seating 
capacity of a theater, the reversed floor principle of planning requires 
no angle of projection greater than ten degrees. 

It has been found that the enlarged screen can be more easily ac- 
commodated in a theater structure if the reversed floor principle of 
planning is applied; and that it would be impossible to install an 
enlarged screen in the present type of theater without incurring a 




LONGITUDINAL- SECTION- 

MOTION- PICTURE THEATRE - 

FIG 1. Longitudinal section of the present type of motion picture theater as 
affected by the use of reversed floor. 



great waste of structure area and inefficiency in seating arrangement, 
resulting from the failure to utilize the areas most valuable for com- 
fortable vision. The difficulty of using an enlarged screen in present 
theaters has already evidenced itself, and is partly delaying its popular 
adoption. Balcony obstructions, and the difficulty of obtaining a 
complete and comfortable view of the higher screen are serious im- 
pediments, which may be overcome by the use of the reversed floor. 

The practicability of applying the reversed floor principle to vari- 
ously sized and proportioned plots of ground has been given special 
attention, the object always being to obtain a maximum number of 
"good" seats within a minimum area. Many different adaptations of 
the reversed floor principle have been devised to fit the peculiar con- 



Feb., 1932] 



UTILIZATION OF DESIRABLE SEATING AREAS 



191 



ditions of various theater projects. The feasibility presents itself of 
placing a part or the whole of a theater auditorium above or below 
the portions of a structure which may be used for other purposes. It 
therefore becomes important to design a theater auditorium so that 
it will not require too much valuable area in the vertical sense. Thus, 
the remaining portions of a structure above or below the theater may 
provide an additional income which, in turn, results in a reduced 
rental for the theater itself. The use of the reversed floor permits 
constructing a theater within a limited height, where it would be im- 
possible to include a theater planned according to present practices. 
A revised building code for the city of New York, affecting theater 
structures, is about to be put into effect. The committee revising 
this code has taken into consideration the possibilities of the re- 
versed floor by providing for its development in the wording of the 




FIG. 2. Chart for determining, in the vertical sense, the desirable seating areas. 

code. The revised code will permit the construction of a theater 
auditorium, having a capacity of more than 600 seats, directly be- 
neath the portions of a building used for other purposes. This is 
already permitted in many other building codes. 

For the purpose of making possible better vision, smaller projec- 
tion angles, and reducing the cubage of the theater structure, a very 
intensive search has been made to ascertain which of the physical 
areas are most valuable as seat locations in relation to the screen, 
the object being to use these areas only for seating arrangements. 
The present system of theater planning necessitates the utilization 
of portions beyond these valuable areas, thus causing large pro- 
jection angles, distorted vision from high balconies, unnecessarily 
large construction costs, and the payment of excessive rentals for 
space which not only has no value to the exhibitor, but which also 
creates conditions highly unsuitable for motion picture exhibition. 



192 BEN SCHLANGER [j. s. M. P. E. 

For these reasons the present method of theater planning results 
in unscientific and uneconomically built structures. If original 
construction and maintenance costs can be reduced, at the same 
time giving to the theater patron the comforts and surroundings due 
him, there is little doubt as to what the effect on the box-office will 
be. 

A chart showing the location of the desirable areas has been de- 
veloped. (Fig. 3.) These areas have been found by determining how 
much above and below, and how far from and how near a fixed screen, 
a spectator may sit, maintaining a comfortable bodily posture, and 
obtain a view of the entire screen. The spectator should be seated 
in such a position that the picture on the screen will appear at a 
level which is most imitative of the level from which natural surround- 
ings are viewed in real life. Still another determining factor in 
locating these valuable areas is that it is more natural to sit low and 
lean slightly backward against the chair to obtain a higher view, 
than it is to sit high and lean forward to look down at a screen below 
the eye level. For these reasons, therefore, the desirable areas of 
seating are limited to levels below the top of the picture. While 
both the reversed orchestra floor slope and the present orchestra floor 
slope come within the desirable areas, the present slope of floor, which 
rises up away from the screen, causes all upper levels of seating to 
come within the areas undesirable for natural and comfortable view- 
ing. The reversed orchestra floor slope has a much smaller pitch 
downward than the present type of floor has upward. The present 
type of floor eats into the valuable areas unnecessarily, while the 
reversed floor hugs the lower region of the desirable areas, leaving the 
remaining valuable areas for additional seating levels. 

A method of adjusting the level of the screen, the levels of the vari- 
ous eye lines, the distance between the eyes and the screen, the slope 
and inclination of the various levels of seating, and the pitches of the 
backs of the seats has been developed, keeping a definite relation 
between all the elements involved. (Fig. 3.) Formulas have been 
evolved to define the position of the screen in relation to the slopes 
required for the orchestra and balcony levels. The shape and size 
of the screen are also a definite part of the calculations. Given the 
shape and size of the screen, and the distance between the screen and 
the nearest seat (which should equal the width of the screen, for per- 
fect horizontal vision), the vertical distance between the level of the 
nearest seat and the level of the bottom of the screen is determined. 



Feb., 1932] UTILIZATION OF DESIRABLE SEATING AREAS 



193 




194 



BEN SCHLANGER 



[J. S. M. P. E. 



This distance determines the slope of the orchestra floor and the 
pitches of the chair backs. The resultant slope is a parabolic curve 
which starts near the screen with a downward pitch, decreasing uni- 
formly until the floor is practically flat at about the twenty -fourth 
row of orchestra seats. 

The slope of the orchestra floor in existing theaters is not sufficient 
to permit seeing the bottom of the screen over the head of the person 
immediately ahead. If the pitch were sufficient for this purpose, it 
would be too great for comfortable walking, and would make it 
difficult to adjust the standards and leg supports of chairs to the slope 
of the floor. The mild pitch of the reversed floor necessary to allow 
full view of the bottom of the screen eliminates these difficulties. 




FIG. 4. Adaptation of reversed floor to a small theater. 

A full, comfortable view of the entire screen can not be obtained until 
the ninth row of the present orchestra floor slope is reached, causing 
severe neck- and eye-strain in about 300 seats in the average theater. 
The reversed floor corrects this condition entirely, allowing a comfort- 
able view from every seat in the orchestra and higher levels. 

The slope of the reversed floor automatically establishes the proper 
pitch for the backs of the seats in every row. This eliminates the 
need for specially adjusted backs, and changes in the standards and 
leg supports of the chairs. All chairs can then be exactly alike in 
every detail of construction. Instead of designing differently con- 
structed seats to fit a floor slope, as is now necessary, the floor is de- 
signed to suit uniform seating. It is just as though the seats were 
placed in an ideal position for viewing the screen, the floor being built 
afterward to support the seats in the proper manner. 

The matter of determining the maximum tilt for chair backs has 
been discussed with the engineering department of a leading seating 



Feb., 1932] UTILIZATION OF DESIRABLE SEATING AREAS 195 

company. A tilt of twenty-seven degrees for the row of seats nearest 
the screen has been suggested. This tilt will equally distribute the 
weight of the body to the seat and back of the chair, keeping the head 
of the spectator in a comfortable position. The tilt diminishes to 
sixteen and two-thirds degrees at about the twenty-fourth row, re- 
maining constant thereafter. The fact that the head assumes a pitch 
of two or three degrees less than the pitch of the body has been taken 
into consideration in the calculations for determining the floor slope. 
The feet of the occupant of a seat are properly supported, and the 
sensation of sliding forward out of the seat, as experienced in the 
present type of orchestra seating, is eliminated, because the angle and 
distance between the floor and the seat remain constant. Due to the 
fact that the same chair can be used unchanged throughout the house, 
a considerable economy can be effected. The "spring edge" seats, 
required where the present type of floor inclines steeply from the 
front of the chair, are costly and unsatisfactory, and can be elimi- 
nated. The reversed floor permits the use of the "box spring" seat, 
which is less costly and more durable. The uneven wearing of the 
seats and backs of chairs is also corrected by equally distributing the 
weight of the body by properly tilting the seats on the reversed floor. 
The results of all these studies and tests have been utilized in pre- 
paring a series of new forms for motion picture theater structures. 
Many variations of the seating arrangement are made possible that 
could not have been arrived at with the present type of floor. The 
chart locating the valuable areas has been used as a basis for designing 
these forms. 

DISCUSSION 

MR. KELLOGG: From a novice's standpoint, I have sometimes thought of how 
I might try to figure out the best arrangement of theater floor and seats. Imagine 
yourself looking toward the audience from the center of the screen; if the solid 
angle which your eyes encompass were filled as compactly as possible with eyes 
and ears, that would be the way you could get the most people in. Thinking 
of it from this standpoint, there are several parts of the possible solid angle, 
within which people might hear and see satisfactorily, that are poorly utilized. 
One is the region below, which the arrangement proposed by Mr. Schlanger is 
primarily aimed to utilize to better advantage; and the other is the part of the 
solid angle occupied by the fronts of the balconies. There is only one way 
to cut out waste space due to the balconies, besides making them as thin as 
possible, and that is, not to have balconies. But that does not furnish good 
space utilization, either. 

From the standpoint of the utilizing angle below, one is confronted with the 



196 BEN SCHLANGER [j. s. M. P. E. 

fact that the part of the audience nearest the stage subtends a disproportionately 
large angle compared with the number of people. I am wondering whether, 
following the idea described in the paper, we might not actually, within the proper 
angle, pack more people into a theater of given dimensions by not trying to be- 
gin too close to the screen. If you drop a little further back you can begin a 
little lower, leaving more height available for balconies. 

There is one question I do not believe was answered in the paper, and that is, 
what is considered the desirable rise per row of seats. To be more specific, if 
a line were drawn from the back of one row of seats to the bottom of the screen, 
how much would that line miss the back of the next row? As a general rule, 
one can probably figure on looking between the heads of the people immediately 
in front, but he will have to look over the heads of those seated two rows ahead 
of him. I should be interested to learn how much allowance is made for this 
factor. 

MR. SCHLANGER : The point about having the first eye line farther away from 
the screen is quite possible, but it depends on how much area in front of the 
screen the theater exhibitor is willing to devote to it. That area is costly. It 
is better practice to place the first seat a little farther away from the screen 
than it is usually placed. 

Referring to Fig. 3, it is seen that a given person sees above and not between 
the heads of the persons in front of him. His sight line passes over the head 
in front, passing to the very bottom of the screen. The orchestra floor in the 
present type of theater is not pitched sufficiently to allow this complete view of 
the screen, therefore making it necessary for people to keep shifting their positions 
in order to see between the heads of the people ahead. 

When constructing the diagram, a line is drawn from the bottom of the screen 
to the top of the head of the first spectator. This gives the eye level for the 
next row. The distance from the eye to the top of the head is four inches. So, 
connect another line from the bottom of the screen to a point four inches higher, 
and we have the sight line of the next row, and so on. 

The pitch of the floor varies from row to row. The first row requires a greater 
pitch than the succeeding one, and so on. On reaching the twenty-fourth row, 
in the reversed floor system, the floor becomes practically flat. If the theater 
were big enough the curve would rise again somewhere about the fiftieth row. 

In the present type of house, the low placing of the screen makes it necessary 
for each succeeding head to be higher, in order to see over the head of the person 
in front. The pitch increases so rapidly that in existing theaters a compromise 
has been effected, and one looks between the heads of those in front, and not 
over their heads, as on the reversed floor. 

MR. PORTER: I should like to know how much tolerance, if any, is allowed 
for variations in height among individuals. Is any allowance provided for such 
variations, or are the angles worked out for an average height? 

MR. SCHLANGER: The average height measured from the floor to the eye level, 
the person sitting in a normally pitched chair on a flat floor, varies from three 
feet seven inches, to four feet; four feet is the dimension for a person six feet 
two inches tall. To allow for a person five feet eleven inches, or six feet, would 
be about right. Therefore, about three feet eleven inches should be assumed. 
That would take care of almost all such conditions. 



Feb., 1932] UTILIZATION OF DESIRABLE SEATING AREAS 197 

MR. RICHARDSON: On what do you base your choice of what are termed 
desirable areas? 

MR. SCHLANGER: The desirable areas are determined by the distance above 
or below the level of the screen at which it is comfortable to view the screen. 
To establish areas of desirability, a line is drawn from a test point to the impor- 
tant focal point on the screen; a line is then drawn, representing the back of the 
spectator, ninety degrees to this line. This will show how much the spectator 
must lean forward or backward. The areas where the spectator does not have 
to lean forward or raise his head are most desirable. The degree of deviation 
from comfortable posture determines the relative value of a given area. 

MR. SPENCE: The suggestion was made that this floor plan would be adapt- 
able to a Trans Lux theater, but I do not think so. The Trans Lux theater was 
designed with a ten-foot head room, in order to be able to move in and out of 
established office buildings. To use this type of floor it would be necessary to 
break through to the cellar. If rear end projection were used, it is possible that 
people walking in the aisle would get into the beam, and the seats would have to 
be placed farther back from the screen. The angle of view is so much wider 
with a rear projection screen, that what would be gained by putting a projection 
room near the box-office, so to speak, would be lost at the front. 

MR. SCHLANGER: The distance from the first row to the screen is not deter- 
mined by the method of projection. It is determined by the size of the picture. 
Because the Trans Lux theater presents a small picture frame on the wall and 
we can get very close to it, does not mean that we can use Trans Lux projection 
and have effective motion picture exhibition. 

As to the ten-foot height mentioned, the head room of the first floor of the 
average building is nearer twelve feet. It is possible to use standard projection 
in such limited heights by employing a reversed floor, thereby requiring only 
half the projection booth area necessary for Trans Lux projection. The saving 
in annual rental for this space on the street level could be used to rent a foot or 
two below the street grade, which would allow the use of a much larger screen 
than is now used in Trans Lux theaters. 

MR. PORTER: What would be the total drop in your floor? 

MR. SCHLANGER: The slope of the reversed floor is considerably less than 
the slope of the present type of floor. 

MR. Fox: Mr. Schlanger showed that the weight was equally distributed be- 
tween the back and seat of the chair. He meant that a certain amount of weight 
was taken from the seat and applied to the back. About fifteen per cent of the 
excessive tilt is transferred to the back from the seat. We have done what the 
airplane man has done, so that the passenger would not fall out of his seat when 
the plane was tilted for landing. 

MR. HICKMAN: I should like to ask if any theaters have been constructed in 
this manner? 

MR. SCHLANGER: Mr. Kinsila, in his book on theaters, stated that a theater 
having a reversed floor was built in Moscow many years ago. Better Theatres 
published an account of the Pathe Theater in Paris, which also has a reversed 
floor. In both cases, there is no evidence of a curved reversed floor. The 
Pathe Theater was constructed in such a manner because of a steep street grade 
condition, and a limited height usable in an existing building. The posture 



198 BEN SCHLANGER 

and sight line problems in this theater were not given any thought, as the uniform 
floor pitch and chair tilts show. 

At present, I am working on a small theater, which is now under construction, 
in which the parabolic reversed floor is used. To my knowledge, this will be the 
first theater built wherein the principles of applied optics and good sitting postures 
are recognized. 



A METHOD OF MEASURING DIRECTLY THE DISTORTION 
IN AUDIO FREQUENCY AMPLIFIER SYSTEMS* 

W. N. TUTTLE** 



Summary. The question of a suitable measure of harmonic distortion is dis- 
cussed. The distortion factor employed is defined as the ratio of the effective value 
of the combined harmonic voltages to the fundamental voltage. A simple rapid 
method of measuring this ratio is described which has several advantages over earlier 
methods. Results are given showing the performance of apparatus for the applica- 
tion of this method to the testing of audio frequency amplifier systems. 

It is customary to rate an amplifier used in the reproduction of 
speech or music on the basis of the amount of undistorted power which 
it is capable of delivering to the loud speakers. 

The distortion produced by the amplifier is the deviation in the 
shape of the electrical wave of the amplifier output from that applied 
at the input terminals. Several types of distortion are evidently 
possible. Frequency distortion takes place when the system does 
not amplify equally all the component frequencies of the input wave. 
Phase distortion occurs when these several component voltage waves 
are shifted in time with respect to one another. Harmonic distortion 
or amplitude distortion is observed when the crests of the voltage 
wave tend to be partly cut off. The first two of these effects, fre- 
quency distortion and phase distortion, depend largely on the funda- 
mental design of the amplifier and are not appreciably affected by 
variations in the magnitude of the input voltage. Harmonic distor- 
tion has the opposite characteristic. It is important when the 
amplifier is operated at low energy levels, but increases rapidly when 
the voltage applied is increased beyond a certain value. Harmonic 
distortion is the factor which limits the useful output of an amplifier. 
It is the measurement of this effect which is to be considered. 

The measure taken for harmonic distortion should, if possible, 
be a measure of the objectionableness to the listener of the distortion 
present. Different types of amplifiers cause different types of distor- 

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. 
** General Radio Co., Cambridge, Mass. 

109 



200 



W. N. TUTTLE 



[J. S. M. P. E. 



tion, so that the quantity measured should be such as to supply a basis 
for the comparison of amplifiers of different fundamental design. 

If a single pure tone is applied to the input terminals of an amplifier 
harmonic distortion will result in the appearance in the output 
voltage wave of various harmonic frequencies of the applied tone. 
Let us call EI the fundamental voltage, and 2 , 3, 4 ... the 
voltages of the several harmonic frequencies. Let us consider the 
ratio 



D 



This is seen to be the ratio of the effective value of the combined 
harmonic voltages to the fundamental voltage. This expression 
gives equal weight to all harmonics having the same value, but as each 



TEST 
VOLTAGE <> _ 




FIG. 1. Simplified functional diagram of the distortion-factor meter. 

harmonic enters as its square, prominent components are considerably 
emphasized. A single component of two volts, for example, is given 
twice the importance of two harmonics, each of one volt. In view of 
the masking effect of one tone by another, the single two- volt compo- 
nent would be expected to be correspondingly more objectionable. 
In view of these considerations and of the fact that it lends itself 
readily to direct measurement, the ratio D has been generally 
adopted 1 ' 2 as the best measure of harmonic distortion. 

The methods of evaluating the distortion factor which have been 
available have been suited primarily to laboratory use. One method 
is that of measuring the separate harmonics in the amplifier output by 
means of a harmonic analyzer, and of computing from these values the 
distortion factor. Similarly, oscillographic records can be analyzed 
approximately by any one of several methods. A less laborious 



Feb., 1932] 



MEASURING DISTORTION 



201 



procedure is that of eliminating the fundamental in a suitably designed 
bridge circuit and measuring the combined harmonics which remain. 3 
This method has one serious disadvantage in that the bridge must be 



TYPE 536 A 

DISTORTION FACTOR METER 
FILTER CHARACTERISTIC 



100 



1000 5000 10000 

FREQUENCY - CYCLES PER SECOND 



500OO 



FIG. 2. Over-all attenuation characteristic of the filter and input resistances. 

carefully balanced, and that slight fluctuations in the test frequency 
may consequently make it difficult to obtain satisfactory results. 
Another disadvantage is that frequencies below the test frequency, 
including power-supply hum, are included with the harmonics. 



INDICATOR Z 
> 500000/1 




FIG. 3. Detailed circuit diagram of the distortion-factor meter. 



Ballantine and Cobb 4 have developed an ingenious null method of ob- 
taining the distortion factor which avoids these difficulties. But 
this method, also, is suited principally to laboratory use. 



202 W. N. TUTTLE [J. S. M. P. E. 

It seemed desirable to develop a rapid method of measuring the 
distortion factor which would require neither bulky apparatus nor the 
services of a skilled operator. The essentials of the method finally 
adopted are shown in Fig. 1. 

The test voltage is applied to a high-pass filter and an attenuator in 
parallel. The output of the filter is proportional to the combined 
harmonic content of the test voltage. The output of the fixed 
attenuator is proportional to the test voltage itself, which is approxi- 




FIG. 4. Panel arrangement of the distortion-factor meter. 

mately equal to the fundamental voltage in cases encountered in 
practice. We can therefore determine the distortion factor by 
comparing the output of the filter with the output of the attenuator. 
In the apparatus constructed, the fixed attenuator is so proportioned 
that when the voltage across the entire voltage divider is equal to the 
filter output voltage, the distortion factor is 30 per cent. A dial 
reading directly distortion factors from zero to 30 per cent is at- 
tached to the voltage divider control. All that is necessary to make a 



Feb., 1932] MEASURING DISTORTION 203 

measurement of the total harmonic content of the test voltage is to 
observe the deflection of the indicator when connected to the output 
of the filter. The indicator is then switched to the output of the 
voltage divider, the setting of which is varied until the same deflection 
is obtained. The voltage divider scale reading then gives the 
distortion factor directly. 

The success of this method evidently depends on the care with 
which the apparatus is designed rather than on the skill of the operator. 
The filter must reduce the amplitude of the fundamentals so that it is 
negligible compared with the harmonics at the lowest distortion 
factor to be measured. It must transmit the harmonics equally and 




FIG. 5. Distortion-factor meter with associated amplifier and square-law 

galvanometer. 

must not act as a generator of harmonics even when large input 
voltages are applied. The measured transmission curve of the filter 
developed for this purpose is given in Fig. 2. 

It is seen that the fundamental (400) cycles are reduced relative to 
the harmonics more than 70 db. All harmonics up to the fifteenth 
are transmitted equally within 0.3 db. The attenuation of the 
fundamental is sufficiently great so that the test frequency can vary 
over a range of more than 50 cycles without affecting the result. It 
is seen that enough attenuation is provided at the lower frequencies to 
eliminate power supply hum. 

The circuit details are shown in Fig. 3. It will be observed that a 
series resistance is placed in the input branch. This keeps the 



204 



W. N. TUTTLE 



[J. S. M. P. E. 



impedance out of which the filter works practically constant and 
makes the calibration independent of the impedance of the voltage 
source. It also results in the impedance at the input terminals being 
high enough (about 175,000 ohms) so that the instrument may be 
connected across practically any element in amplifier circuits without 
causing appreciable disturbance. Due to the input impedance 
characteristic of the filter alone, the series resistance reduces the 
harmonic content of the voltage across the voltage divider to such 
an extent relative to the fundamental that no correction need be 
applied. The "L" network RiR 2 may be switched into the circuit 
to magnify the scale of the dial by ten when measuring small distor- 
tion factors. 



24 



X 



X 



I 2 3 4 

POWER OUTPUT - WATTS 

FIG. 6. Curves of distortion factor vs. power output for a laboratory 
amplifier employing a single 245-type tube, for two values of load resistance. 

The apparatus is simple enough so that a compact mechanical 
arrangement is possible. Fig. 4 shows the panel arrangement. 

The indicator must have an input impedance high compared with 
the filter impedance. As high sensitivity is also required in many 
measurements, it is convenient to employ an amplifier in conjunction 
with an a-c. voltmeter. To keep the impedance high, no input 
transformer should be used preceding the first amplifier tube. The 
voltmeter should be of a type that will indicate the effective value of a 
composite voltage. Thermocouple instruments have this property 
but are sluggish in action and are easily burned out. Vacuum 



Feb., 1932] 



MEASURING DISTORTION 



205 



tube voltmeters of the square-law type are satisfactory. A rectifier 
type 2-C galvanometer has been developed for use with the distortion 
factor meter which has a characteristic closely approximating a 
square law and which combines ruggedness with high sensitivity. 
The distortion-factor meter, together with the associated amplifier 
and the a-c. galvanometer, is shown in Fig. 5. 




1000 



2000 3000 4000 5000 
LOAD RESISTANCE - OHMS 



6000 7000 



FIG. 7. Curves of distortion factor vs. load resistance for three values of 
output power for the amplifier of Fig. 6. (The ratio RL/RP of load resistance 
to plate resistance is indicated for reference.) 



Curves obtained with this apparatus in testing a laboratory ampli- 
fier unit are shown in Figs. 6 and 7. Jn obtaining the data of Fig. 6, 
the load resistance was held constant as the input was varied. Ob- 
servations were made of the load voltage and distortion factor as the 
amplifier input voltage was increased. The distortion factor is 



206 W. N. TUTTLE 

plotted against the computed power output for two values of load 
resistance. 

Fig. 7 gives the data obtained by simultaneously varying the load 
resistance and the input voltage to keep the output power constant. 
It is interesting to note the manner in which the optimum load 
resistance varies with the allowable distortion factor. 

These results support the conclusion that the distortion factor 
which has been defined is the logical index of the performance of an 
amplifier. The apparatus developed for directly measuring this 
quantity may be conveniently used in making the simultaneous 
measurements of power output and total harmonic content which 
are necessary in obtaining a definite output rating for an amplifier. 
The apparatus is also suited to checking the operating condition of an 
amplifier installation. A single measurement of the distortion factor 
at the rated power output indicates definitely whether or not the 
system is functioning properly. 

In view of its adaptability to the testing of amplifier systems, it is 
hoped that the distort ion -factor meter will prove useful to the motion 
picture industry in maintaining definite standards of amplifier 
performance. 

RFFERENCES 

1 BALLANTINE, S.: "Detection at High Signal Voltages," Proc. I. R. E., 17 
(July, 1929), p. 1153. 

2 WOLFF, I.: "The Alternating-Current Bridge as a Harmonic Analyser," 
/. Opt. Soc. Amer. and Rev. Sci. Instr., 15 (September, 1927), p. 163. 

3 BELFILS, G.: "Measuring the Residue of Voltage Curves with a Distortion 
Factor Meter," Rev. gen. de I'electricite, 19 (April 3, 1926), p. 523. 

4 BALLANTINE, S., AND COBB, H. L.: "Power Output Characteristics of the 
Pentade," Proc., I. R. E., 18 (March, 1930), p. 450. 



DIRECTIONAL EFFECTS IN CONTINUOUS FILM 
PROCESSING* 

J. CRABTREE** 

Summary. Continuous motion picture film developing machines set up a uni- 
directional current of developer as a consequence of the motion of the film unless 
vigorous measures are adapted to counteract it. This current is shown to cause dis- 
tortions of the H & D characteristics of the photographic material, when the usual 
type of sensitometer strip is used for control of the development. The effect of the 
geometry of the sensitometer strip on the extent of "direction effect" is discussed. 

The difficulty of obtaining equal degrees of development of photo- 
graphic images over even relatively small areas of any given emulsion 
layer is well known to workers with these materials. The necessity 
for constant and vigorous agitation of the developing solution across 
the light-exposed area to be developed becomes evident even to the 
amateur. 

The reason for the desirability of vigorous agitation during the 
development of a photographic image is the necessity for removal from 
the neighborhood of that image, of the products of chemical reaction, 
chiefly bromides and developer oxidation products, which diffuse 
out from the emulsion layer, and which, apart from locally reducing 
the developing power of the solution by exhaustion, have in them- 
selves an actual depressing effect on density. These reaction prod- 
ucts, being of greater specific gravity than the original developing 
solution, probably set up slight local eddy currents during static 
development which give rise to unevenness in the degree of develop- 
ment from point to point, even though the exposure to light is 
uniform. The unevenness is relatively the more pronounced, the 
less the degree of development, tending to disappear as gamma 
infinity is approached. The well-known "Eberhardt effect" or 
"Mackie Line" is occasioned in this manner. These designations 
refer to the light halo appearing around an area of high density, in a 

* Presented in the Symposium on Laboratory Practices at the Spring, 1931, 
Meeting at Hollywood, Calif. 

** Bell Telephone Laboratories, New York, N. Y. 

207 



208 J. CRABTREE [J. S. M. P. E. 

field of lower density. The halo results from the reduced rate of 
development of the area of lower density adjacent to the area of 
higher density, consequent on the diffusion outward from this high 
density area of concentrated reaction products of development. A 
study of this effect was recently made by Walenkov 1 who gives a 
bibliography of the literature. 

Before the introduction of machine development for motion picture 
film, this material was processed almost entirely by the rack and tank 
system, in which the film was wound on large square or rectangular 
racks which were immersed in deep tanks of suitable shape. The 
unevenness of development obtained along the length of film handled 
in this manner was well recognized, and has been thoroughly discussed 
by J. I. Crab tree and C. E. Ives. 2 The method gave locally increased 
development in the region of the rack-ends and cross-bars, due to 




FIG. 1. Strip of dyed film showing effect of directional currents. 

production of eddy currents from temperature differences as well as 
from differences in specific gravity of reaction products, unless agita- 
tion of the developer was obtained by motion of the rack, which was 
not commercially practicable. Pictures developed in this manner 
showed periodic light and dark bands when the resulting print was 
projected upon the screen, as well as Eberhardt effects when de- 
velopment was restricted to allow for control in contrast of the 
picture. 

In recent years an already growing tendency to change from rack 
and tank development to so-called continuous development by 
machine was stimulated by the addition of sound to the picture, 
so that, as a result, almost all such film is now processed by machines 
in which the film is mechanically propelled at a uniform rate through 
horizontal trays or deep vertical tanks containing the developing solu- 
tion in circulation. This method of processing has resulted in such an 



Feb., 1932] 



DIRECTIONAL EFFECTS IN PROCESSING 



209 



apparent uniformity of product that certain local effects have been 
lost sight of. 

However, if the developing bath of a continuous film processing 
machine is observed during operation, it will be seen that the onward 
movement of the strands of film sets up a current of developer in the 
direction of the film movement, but which, relative to the film itself, 
is in a direction opposite to that of the film travel. 

The result of this is shown clearly by Fig. 1, a photograph of a 
section of film which was first dyed red over a small area, then passed 



HIGH DENSITY LEADING 
LOW DENSITY LEADING 




FIG. 2. Typical "direction pair" H& D curves; machine development. 



through the developing and fixing trays of an Erbograph machine at 
sixty feet per minute in the direction shown by the arrow. It will be 
seen that as the color diffused out from the dyed area it was carried 
backward along the film surface and partly absorbed by it. 

During the development of a photographic image, the products of 
the chemical reaction taking place diffuse out from the gelatin layer 
much as the solution of dye did in the manner shown in Fig. 1. 
Therefore, when developing motion picture film in such a continuous 
processing machine as the above, the products of reaction of the 



210 



J. CRABTREE 



[J. S. M. p. E. 



development of any image must flow across the images immediately 
following it. Since, as was previously mentioned, these products of 
reaction have a restraining effect on development, processing by this 
means will result in a variation in degree of development from 
point to point on the film depending upon the concentration of the 
reaction products at those points, which in turn depends upon the 
magnitude of the density of the image area just ahead. 

Let us now consider what happens in the development of a sensi- 



3.2 



2.8 



2.4 



2.0 



1.6 



0.8 



HIGH EXPOSURE LEADING 
LOW EXPOSURE LEADING 
BRUSH DEVELOPMENT (THIS 
CURVE IS SPACED FROM THE 
MACHINE PAIR FOR EASY 

COMPARISON) r 




LOG RELATIVE EXPOSURE 



FIG. 3. "Direction pairs" H & D curves; machine develop- 
ment, showing comparison with brush development. 



tometer exposure. This strip of exposed film bears a series of latent 
images increasing progressively in magnitude from step to step and 
which on development will result in a series of density areas, ab, 
in Fig. 2, represents diagrammatically such an exposed film, the 
shaded area being the image portion, and the clear area representing 
unchanged silver halide. If now, this strip is moved steadily through 
the developer in direction AB (that is, B meeting the developer first), 
the reaction products of development from the first step will, by 



Feb., 1932] 



DIRECTIONAL EFFECTS IN PROCESSING 



211 



virtue of the motion of the film, flow over and successively affect 
the other steps. The initial reaction products will be reenforced by 
the reaction products from succeeding steps, causing a progressive 
weakening of the developing effect as the end A is approached. 

With the strip traveling in the opposite direction (A leading) 
the developer around end A will be little affected by reaction products 
since these will be small in amount and hence nearly full development 
will be obtained. As B is approached, however, the products ac- 
cumulate from the gradually increasing densities so that there is a 



2.8 
2.4 
2.0 

1.6 
> 
H 
</) 

a.i. 

0.8 
0.4 










GH DENSITY UP 
DW DENSITY UP 












B 1 




L 












^ 


S* 


B 


























/s 


,' 




























/* 


'' 




























x 


// 






























// 






























/ 


'/ 






























^' 








y 


-B' ! 
















^ 








^ 




*^~ 


B 
















7, 


^' ' 






S?- 


S' 






















z 




^ 


<S' 
























?' 


A 


^ 


x 






















y^ 


'g 


^ 
























^ 


^ 


^ 























O 0.4 0.8 1.2 1.6 2.0 2.4 2.8 32 
LOG RELATIVE EXPOSURE 

FIG. 4. "Direction pairs" H & D curves; rack tank and development. 



decided loss of density in the region of B. The characteristic H & D 
curves plotted from the readings of density obtained in the two cases 
cited will be found to have the general shapes shown in Fig. 2 where 
A'B' is that resulting from progression through the developer with 
the high exposure leading; while AB is obtained when development is 
carried on with the low exposure leading. 

The general effect is to straighten out the shoulder in A'B' and to 
depress it somewhat in AB. The effect on gamma is slight but there 
is an appreciable difference in the estimate of latitude (straight line 



212 



J. CRABTREE 



[T. S. M. P. E. 



portion) and inertia (intercept on log R axis) in the two cases. That 
neither curve is true in form is shown in Fig. 3, in which similar 
pairs of curves are shown at two gammas with the corresponding 
curve obtained by "brush" development. This "brush" method is 
one of those generally used in precise sensitometry and consists in 
securing a very thorough removal of reaction products from the film 
surface by passing a soft brush rapidly backward and forward across 
it during development. 

A similar effect was found to result from rack and tank develop- 
ment where the manipulation of the rack was such as to result in a 




FIG. 5. Perspective of Erbograph type developing tray. 

moderate amount of agitation of the developer. (See Fig. 4.) With 
this method, curves of type A'B' resulted from those cases where 
the sensitometer strip was developed with the high exposure end of 
the test strip uppermost, while AB resulted from development with 
the lowest exposure uppermost. Since in this case the reaction 
products which are of higher specific gravity than the original de- 
veloper tend to generate a downward current, the same explanation 
applies to the occurrence of the two types of curve as in the case of 
machine development. 

Since the development of variable density sound records in con- 



Feb., 1932] 



DIRECTIONAL EFFECTS IN PROCESSING 



213 



tinuous machines is usually controlled by using a sensitometer strip 
in some form, an inquiry was made to determine what sensitometric 
troubles might be encountered in practice from this "directional 
effect," as it will be called in this paper, and to determine the best 
type and manner of use of the sensitometer strip in machine develop- 
ment. 

In the processing of variable density records the usual procedure is 
to develop the sound negative to an approximate gamma of 0.6 in' a 
developer of the D-76 borax type, and to develop the print in a D-16 



2.8 



2.4 




0.4 



2 4 6 8 10 

DEVELOPMENT TIME IN MINUTES 

FIG. 6. Time-gamma curves for the negative and positive de- 
veloper used. 

type of bath to a gamma of 1.80, or higher. Attention was mainly 
confined to a study of these two types of developer at gammas in the 
region of those just mentioned. Also, the directional effect in the 
particular machine used has been found to be present for a variety of 
commercial types of film, although in this study we used only standard 
positive film. 

The machines used in the experiments to be described were of the 
Erbograph type, in which the film is stranded horizontally around 
drive rolls in 15 loops of 13 feet each in a horizontal tray of 50 gallons 



214 



J. CRABTREE 



capacity, as shown diagrammatically in Fig. 5. Circulation was by 
gravitational feed, and overflow to the return pumps was at the 
rate of 10 gallons per minute. The film speed may be adjusted over a 
range of from 10 to 100 feet per minute. 




A' 
TIME 



I IMC 

FIG. 7. Typical time-gamma curve of a photographic developer 
The formulas of the developers employed in this work are: 



Negative 


Positive 




Elon 


2 grams 


Elon 


0.3 gram 


Hydroquinone 


5 grams 


Hydroquinone 


6.0 grams 


Sodium sulphite 


100 grams 


Sodium sulphite 


37.0 grams 


(anhydrous) 




(anhydrous) 




Borax 


8 grams 


Sodium carbonate 


12 . 5 grams 






(anhydrous) 




Boric acid 


8 grams 


Potassium bromide 


0.9 gram 


Water to 


1 liter 


Water to 


1 liter 



Their time-gamma curves for machine development at 67 F. 
are given in Fig. 6. 

The sensitometer exposures were made in a variable intensity 
sensitometer using photographic step tablets. Different dimensions 
of tablet were used, as will be explained later. Where not otherwise 




0.4 





0.4 0.8 1.2 1.6 2.0 2.4 2.6 3.2 

LOG RELATIVE EXPOSURE 

FIG 8. Effect of gamma on "directional effect;" negative 
developer. 



3.0 
3.2 
2.8 

2.4 

>.2.0 
2 

... 

1.2 

ae 

0.4 



HIGH DENSITY LEADING 
LOW DENSITY LEADING 


















/ 












x 





















Jr 










"/ 




















7 








s> 


^*- 


^ ' 




















1 






f 








x 






















// 






^^ 


^ - 


















^/ 




/ f 


/ 




^ 




















$J 






// 




// 




















"]_ 




/ 


/ 


// 
























P 


* 


// 


/ t 


f 






















// 




1 

/A, 


r >^J 


'/ 
























I 


r 


''y// 
























> 




/ 


z 


























t 




/ 


^ / 


























































^ 


'/'' 




























/. 


*'/' 


' 
























-- ^- 


^ 


^ 





























0.4 0. \2. 1.6 2.0 2.4 ZA 
LOG RELATIVE EXPOSURE 

FIG. 9. Effect of gamma on "directional effect;" positive 
developer. 



216 



J. CRABTREE 



[J. S. M. p. E. 



specified, the length of each step was 5 /i 6 inch exposed across the full 
width of the 35-mm. motion picture film. 

Pairs of sensitometer strips exposed in exactly the same manner 
were passed through the machine, one with the lightest exposure, 
the other with the heaviest exposure leading and will be referred to 
as "directional pairs." The differences between the two H & D 



3.2 



2.8 



2.4 



2.0 



1.6 



5 ,.2 



0.8 



0.4 



0.4 




POSITIVE 



NEGATIVE 



LOG RELATIVE EXPOSURE 

FIG. 1Q. H & D curves showing effect of circulation of developer in machine 

development. 

curves resulting are considered as an approximate measure of the 
"directional effect." 

EFFECT OF GAMMA ON DIRECTIONAL EFFECT 

As is well known, the effect of an increase of circulation during 
development is to decrease the time required to attain a certain degree 
of development. This is a result of the more rapid removal of reac- 
tion products from the emulsion surface of the film. We may there- 
fore, for our present purpose, consider the effect of the accumulation of 



Feb., 1932] 



DIRECTIONAL EFFECTS IN PROCESSING 



217 



reaction products at any point to be equivalent to a loss of effective 
time of development. Reference to a typical time-gamma curve in 
Fig. 7 indicates that this will be more important at low gammas, 
since a given interval of time has, in that region of the curve, more 
effect on gamma or density. 

Since the "directional effect" is considered to be a result of local 



CROSS BARS CARRYING FILM WIPERS 




TRAY 



PLAN 



CROSS BARS 







WIPERS 



r luivi 

* 



j] ITI m [J] DO DD [J] [B_ 



FIG. 11. Plan and elevation of device for eliminating "directional effect" 
in machine development. 

accumulations of reaction products, its degree might well be expected 
to vary with gamma. 

Figs. 8 and 9 show three "direction pairs" at low, intermediate, 
and high gammas in the negative and positive developers, respec- 
tively. From these curves it will be seen that directional effect is 
present at all gammas likely to be used in negative development but 
that it tends to disappear at higher gammas with the positive bath. 
This is consistent with the remarks on the slope of the respective 
time-gamma curves (Fig. 8) at the particular gammas used. 



218 



J. CRABTREE 



[J. S. M. p. E. 



It should be mentioned here that, as the effect has been found 
to be less pronounced in the positive than in the negative developer, 
the inquiry was, in many cases, restricted to a consideration of 
negative development only. 

INFLUENCE OF DEVELOPER CIRCULATION ON "DIRECTIONAL EFFECT" 

The fact that high gammas (i. e., low film speed) in the negative 
developer showed as much "directional effect" as low gammas (high 




LOG RELATIVE EXPOSURE 

FIG. 12. Typical curves obtained without squeegee device. 

film speed) in the same developer indicates that the general circula- 
tion provided by the pumps to the developer had but little effect in 
breaking up the "direction current." This was further confirmed by 
tests in which for one case no gravity, and pump-return, circulation of 
developer was used; while, for the other, all the circulation which the 
system was capable of giving (ten gallons per minute) was used. 

Fig. 10 shows the curves applicable to the two cases for each 
developer at a film speed of 60 feet per minute. No difference is 



Feb., 1932] 



DIRECTIONAL EFFECTS IN PROCESSING 



219 



apparent that could be considered to be in favor of the circulating 
developer. 

From these results it is evident that additional means of developer 
agitation must be provided if "directional effect" is to be avoided. 
This could be achieved by agitation by such means as paddle, pro- 
peller, jets, or by injection of air. 

It was found, however, that the directional current could be broken 




FIG. 13. 



LOG RELATIVE EXPOSURE 

Typical curves obtained with squeegee device. 



up by diverting it from the surface of the film at frequent intervals 
by the use of a squeegee device. The contrivance used is shown in 
Fig. 11 and consists of a series of stationary rubber squeegees in- 
stalled in the developing tray. Each squeegee is about 6 inches apart 
and set at an angle of 45 degrees to the longitudinal axis of the film. 
As the film passes each squeegee the developer in contact with it is 
diverted sideward into the surrounding mass of developer and is 
so replaced by fresh solution. This is found effectively to prevent 



220 



J. CRABTREE 



[J. S. M. P. E. 



any setting up of continuous currents in the direction of the movement 
of the film. 

Fig. 12 shows curves for negative development at two different 



1.6 



1.2 







































-HIGH DENSITY LEADING 
LOW DENSITY LEADING 










/' 


' 








/ 








^ 


^ 


"** 






^ 


^ 




















s 


s 






s 


/'' 
















FRESH .s 


''' 








s/ 


'EXHAUSTED 












/ 


<^ 


'' 






/ 


s 




















/, 


S^ s 
s 








^ 


,' 
















^^ 


^ 


^"' 






__ -r-e 


^ 


&* 





















LOG RELATIVE EXPOSURE 



FIG. 14. Relation between condition of developer and 
"directional effect." 

gammas, obtained without, and Fig. 13 similar exposures developed 
with such an appliance. It will be observed that the "directional 
effect" has been almost entirely eliminated by the use of this device. 



2.4 



2.0 



1.2 



0.8 



0.4 



HIGH DENSITY LEADING 
LOW DENSITY LEADING 



LOG RELATIVE EXPOSURE 

FIG. 15. Relation between bromide content of developer and 
"directional effect." 

CONDITION OF THE DEVELOPER 

Since the distortion of the H & D curves produced by directional 
currents is caused by the presence of reaction products, it was thought 



Feb., 1932] 



DIRECTIONAL EFFECTS IN PROCESSING 



221 



that if such reaction products were already present in the developer, 
as in the case of a partially exhausted bath, the effect should perhaps 
be relatively less than with a fresh solution. Fig. 14 shows the results 
of such a test and gives "direction pairs" of curves from strips de- 
veloped in fresh negative developer and in the same developer after 
exhaustion to a degree beyond that normally used in practice. The 
improvement shown by the use of the exhausted developer is com- 
paratively slight. 

A similar test was conducted in which the effect of the addition 
bromides to the negative developer was studied. Fig. 15 shows 
curves of "direction pairs" developed in a fresh bath (^4) and also in 
the same bath to which potassium bromide had been added (B). 




LOG RELATIVE EXPOSURE 

FIG. 16. Directional distortion effects in two types of developing machines; 
negative development. 

Little improvement was manifested even though the bromide con- 
centration was for some tests considerably in excess of any prac- 
ticable figure. 

MACHINE DESIGN 

The cause of the "directional effect" is such that its presence and 
amount must depend upon the design of the particular processing 
machine employed. A series of "direction pairs" of strips from step 
tablets of different dimensions were, therefore, processed in another 
laboratory where a vertical-tank type of machine was available. 
A comparison of results obtained with the horizontal tray type used 
in our laboratory is shown in Figs. 16 and 17. The indications are 



222 



J. CRABTREE 



[J. S. M. P. E. 



that the directional effect is somewhat less in the vertical-tank type. 
No information was available as to developers used. It would 
seem reasonable, however, that in the vertical type gravity will assist 
in removal of the reaction products since they are of higher specific 
gravity than the fresh developer. 



SENSITOMETER STRIP DESIGN 



The trailing of reaction products from any given area will influence 
following images only for a certain linear distance for a given density 
magnitude. It would therefore follow that the longitudinal dimen- 




LOG RELATIVE EXPOSURE 



FIG. 17. 



Directional distortion effects in two types of developing machines; 
positive developer. 



sions of a sensitometer exposure strip will have a bearing on the 
differences in the H & D characteristics shown by a "directional pair." 
Also, since a narrow trail would have a better chance of diffusion 
into the surrounding mass than a wide one, the width of the exposure 
should also have an influence. Other factors of importance are the 
density interval between steps and the length of the toe and shoulder 
portions of the curve, respectively. These various factors have been 
examined with the following results. 

(a) Effect of Length of Step. Step tablets were obtained having 
step widths in the longitudinal direction of film travel of 3 /4, Vie, 
Vie, 1 /s, and Vie inch. The 3 /4 and Vie inch tablets were of different 



Feb., 1932] 



DIRECTIONAL EFFECTS IN PROCESSING 



223 



origin and had different density intervals from the remainder but the 
1 /s and Vie inch were identical in origin with the 3 /ie inch tablet, 
having been constructed from a portion of it by dissection and re- 



1.2 
0.8 
0.4 

1.2 

0.8 
0.4 

































































































^ 












" 






HIGH DENSITY LEADING 








'* 






*** 


** 


^*- 








^ 


<* 












LOW DENSITY LEADING 








^ 


^ 










XT 


^ " 




































^^ 


#^ J 




\ 


,. 




^ 


^ 






































^ 


: ' 










*?' 






































<^ 


^ 








x 


{?' 




































^j^l 


^r^ 


^ 








- 


** 


















































































































































































































x _^ 














,' 














^ 


"* 




















^ 


X** 












^ 














^^ 




^. 














. 




X 


-^ 








1 




^ 


^" 









L > <1 




^ 


r^ 5 




















%> 


x^ 












^ 


^" 














^ 






















X 


Xx 

X 












x^ 


^x 










^ 


^V 


" 






















/I 


f ' 










^ 


X X 










^x 


-^- 


' 




















^ 


r^ 


- 










^ 


^ 








r 


^t 




"' 

























LOG RELATIVE EXPOSURE 



FIG. 18. Influence of step width on "directional effect; 
negative development. 



3.2 
2.6 

2.4 
2.0 

S'-6 


1.2 
0.8 
0.4 





1 




\ \ \ 


\ 







































HIGH DENSITY LEADING 
LOW DENSITY LEADING 








/ 




f 










/ 




/ 






* 




^ 









7 






ft 


^ 


* 


S 


^~~ 


s 


** 


























/ 


' 






/x- 


^ 




jt 






/f* 




/ 




























^ 






A 


/ 






/' 






/,' 




A 


' 


























/f 








//' 






// 






// 






// 


























/ 


/ 






/' 






/ 


/ 




^ 


V 




/ 




























// 






t 


\' 






// 






// 






// 



























r 




* 


>/ 


/ 




^ 


ft 




-^ 


'/ 




J/S 


i/f 


























ty 


























// 








// 






/t 






/; 






/" 




























^ 






/, 






t 


// 






// 




I 


[ 


























I 








tt 






// 






^ 






/,' 


























/ 


f 






/> 






f 


' 






f 




j 


































A 








fi 






/ 






J 


























S 






<* 


t 






</' 






^ 






^ 
























^ 


<** 




^f^ 


X 


i* 




^ 


' 






xl 




^ 


S 

























LOG RELATIVE EXPOSURE 

FIG. 19. Influence of step width on "directional effect;" positive 
development. 

assembly. "Direction pairs" of curves are shown in Fig. 18 for 
negative and in Fig. 19 for positive development. They are spaced 
along the log relative E axis for convenience of comparison. Gamma 



224 



J. CRABTREE 



[J. a M. P. E. 



relations do not hold, since development was performed on differ- 
ent occasions, except for the 3 /ie, l /s, and Vie inch curves which were 
developed together. It will be seen that the directional effect is 
more evident in negative development than in positive, and that 
in the negative the effect increases as the width of the step decreases ; 
and that in the 3 /ie, 1 /s, and Vie inch curves in both negative and 
positive the gamma tends to rise as the step width decreases. The 
results obtained in the negative developer appear to be reasonable 
in that the farther away the center of one step is from the center of the 
preceding step, the less the density at the center of the former will be 
affected by the latter, and vice versa. Also, as the longitudinal dimen- 



HIGH DENSITY LEADING 

LOW DENSITY LEADING 




FIG. 20. 



LOG RELATIVE EXPOSURE 

Effect of track width on "directional effect.' 



sion of the sensitometer exposure is reduced, the less will be the general 
dilution of the supernatant developer by the total mass of reaction 
products, and so the gamma reached will be higher. 

(b) Width of Track. Limiting the transverse dimension of the 
sensitometer exposure was found not to show any noticeable diminu- 
tion of "directional effect" so long as the exposure was confined to 
the center of the film. However, limiting the width of the exposure 
to sound track dimensions and its position to that of the sound track 
resulted in a diminished effect compared with the full width exposure. 
The directional effect was very considerably reduced with a 3 / 4 
inch step at positive gammas (Fig. 20) made under these conditions. 



Feb., 1932] 



DIRECTIONAL EFFECTS IN PROCESSING 



225 



(c) The Density Interval between the Steps of the " Sensitometer 
Tablet." The magnitude of the difference in exposure from step to 
step of the sensitometer tablet may be expected to have a bearing 
on the degree of the distortion of the characteristic curve from the 
exposed strips, since the depression of density, by the "directional 
effect," of any given step will depend on the magnitude of the den- 
sity of the step preceding it. A comparison was therefore made be- 
tween tablets having the same exposure range but in which one tablet 
had twice as many steps as the other, and hence but half the density 
interval. 



H 0.8 

z 



0.4 































, 







RANGE = 2.93 
AVERAGE INTERVAL=O.I3 










^f 


^ 


^ 








*$ 


*?* 


























<< 


**' 
























. 


<& 


" 
























^ 


<< 


" 
























^ 


^ 


' 



























1.2 

H0.8 
| 

s 

0.4 





























s 


s 
s 






RANGE =2.97 
AVERAGE INTERVAL^ 0.25 








^ 


^ 




. 








^ 


*>^ 


























^ 


^' 


























st 


x<- 


























^ 


s^z 
























^* 


-^ 


^~* 





























0.4 



O.8 L2 1.6 2.0 2.4 

LOG RELATIVE EXPOSURE 



2.8 



3.2 



FIG. 21. Effect of sensitometer exposure interval on 
"directional effect." 



The curves are shown in Fig. 21 and indicate that decreasing the 
number of steps in the tablet, while not affecting the degree of separa- 
tion between curves of a "pair," shows greater distortion at the 
shoulder and so is less desirable. 

(d) Effect of the Exposure Range of the Sensitometer Tablet. The 
range of exposure in the step tablet type of sensitometer depends on 
the density range of the tablet, and upon this and upon the time and 
intensity of the exposure applied will depend how many of the 
resulting readings of density will fall on the toe and shoulder portions 
of the curve. 



226 



J. CRABTREE 



[J. S. M. P. E. 



1.2 
0.8 
0.4 

n 


















LOW DENSITY LEADING 


































^ " 


^^^ 


^- ' 


^ 


^ - 

^^^ 


P^* 




,**- 


p^ 


-^ 


^ 


^ 




















p. ' 


( ^-* 


K J 


X-* 


"^ 


x^ 


x 


x 


X 


x" 


x 


^* 


























^ 


^ 


x^ 




x^ 


X 


^ 


X 


X 






























^ 




X 




x^ 




x j 







































x 


x 


x 


^ 




































x 


x 


. 


x 








































^^- 


X* 































_ 
b> 



1.2 
>- 

^ : - 

V) 

50.8 

Q 

0.4 



L7 



HIGH DENSITY LEADING 






x 



LOG RELATIVE EXPOSURE 

FIG. 22. Effect of restricted exposure range in sensitometer; absence of toe. 



1.6 



1.2 



Z0.8 



0.4 



























1 1 


















, 




















HIGH DENSITY LEADING 








S 


/ 


^x 


x 


x 


r 
































/ 


x 


' 


x 


x 


x 


,^- 
































/ 




7 


X 


x 


x 


x 


x 
































/ 


X 


x 


x 
x 


x^ 


x 


x 


x 


' 


























X 


X 


s*^ 

/ 


x 

,/ 


x 


x; 


x 


x 


X 


x 


"^ 




























x x 


x 


x 


x 


x 


x 


x 


x 


x 




























g 


<^ 


^ 


-^ 


^^- 


^ 


^-*^ 


?*" 


^ 


^ 


x- 





























1.2 

V 

t- 
<n 

Z0.8 



0.4 
























1 1 1 


1 




































LOW DENSITY LEADING 












^- 


_ > 






































^ 


^ 


^> 


x 






































X 


x 


;x 


x; 


x 


































X 


x 


x 


x 


x^ 


x 


X 
































^ 


x 


X 




x 


^x 


x 
X 


x 
X 


x 






























x 


x^ 






x 




x^ 


X 


X 




























' 


^j 


^^ 


^; 


^ 


^ 


^ 


^ 


<^ 


^ 































LOG RELATIVE EXPOSURE 

FIG. 23. Effect of restricted exposure range in sensitometer; absence 

of shoulder. 



Feb., 1932] 



DIRECTIONAL EFFECTS IN PROCESSING 



227 



It was found that varying these factors resulted in extraordinary 
distortions of the characteristic. Fig. 22 shows curves derived from a 
series of exposures made through a sensitometer tablet in which the 
density steps were progressively masked off from the toe end of the 
curve while in Fig. 23 is shown a similar series in which the steps were 
removed from the shoulder end. The time and intensity of the 
exposure incident on the tablet were constant throughout to ensure 
that no gamma differences could accrue from differences in the 
reciprocity relation. In this case the "direction pairs" are separated 
into two groups, (a) being the group having the high exposure leading 
and (b) the group having the low exposure leading. The curves 



HIGH EXPOSURE LEADING 
LOW EXPOSURE LEADING 



0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 

LOG RELATIVE EXPOSURE 

FIG. 24. Example of errors resulting from "directional effect." 



are spaced along the log relative exposure axis for ease of comparison 
of shapes. They show that unless a full toe and several points on the 
shoulder are obtained in the sensitometer exposure an erroneous 
impression may be gained from the curve resulting from machine 
development, not only of the characteristic, but also of the gamma 
obtained. This is by reason of the fact that whenever the first ex- 
posure to meet the developer happens to be on or near the straight 
line portion, the abnormal increase in density of the first few steps 
will alter the angle of the straight line drawn through the points. 
Under these conditions the effect is to raise the apparent gamma when 
the shoulder end of the straight line meets the developer first and 
to depress it when the toe end leads through the bath. 



228 



J. CRABTREE 



[J. S. M. p. E. 



The conclusion to be drawn is that the sensitometer exposure 
should be arranged to give a full toe and shoulder, and that where 
the range of the sensitometer does not permit this, a full toe should 
be obtained and the strip developed with the toe end leading. 

A typical example of the errors into which one may be led by lack of 
consideration of "directional effect" is that of a particular study of 
an alleged change of gamma with printer point. When a sensitome- 
ter strip is exposed in the printer a full curve will usually be obtained 
at the highest printer point, but as the printer point is decreased, the 
shoulder part, then the upper portion of the straight line, disap- 



1.2 



H0.8 

















8 
>^ 




























/ 


sr 


^X^l 


^ 
























^_ 










^> 


















> 


/ 


DIRECTION THROUGH 
DEVELOPER 

EFFECTIVE y=0.84 

Jo STEP 





\> 












/ 


/ 






\ 


N 








/ 


/ 










N s^ 


>\ 






A 
















*C 






0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 

LOG RELATIVE EXPOSURE 

FIG. 25. "Directional effect" in a simulated sound record. 

pears. The resulting strips are now analogous to the case of the 
shoulder cut-off shown in Fig. 23. If the strips should happen to be 
processed toe end leading, no change in gamma will be noticeable in 
the curves; but should the high exposure end lead, there will be 
gamma distortion at the lower printer points. 

A further case in point relating to field practice is illustrated in 
Fig. 24 which shows a directional pair of H & D curves derived from 
exposures to positive film on a particular time-scale sensitometer. In 
this case an entirely false determination of both gamma and character- 



Feb., 1932] 



DIRECTIONAL EFFECTS IN PROCESSING 



229 



istic would be obtained by this sensitometer from a strip processed 
with its high exposure end leading through the developing bath. 

INFLUENCE OF DIRECTIONAL EFFECT ON THE SOUND RECORD 

A frequency cycle of a variable density sound record consists of a 
series of gradations of density arranged much like pairs of minute 
sensitometer strips with their high densities abutting. It is reason- 
able, therefore, to conclude that "directional effect " in a processing 
machine will distort the recorded sound wave. In a properly ex- 
posed frequency record there should, however, be no shoulder densi- 
ties such as are met with in a sensitometer strip. Exposures from a 



1.2 
1.0 
0.8 
0.6 
0.4 
0.2 

0.6 
OA 

0.2 


















HIGH DENSITY LEADING 
LOW DENSITY LEADING 
























































^N 




[ 


)IRECTION THROUG 
DEVELOPER 


H 
























.^. 


**' 






^^ 


^ 




^ 


<S 


^ 


^^ 














x^' 


*> 












^ 


^ 






** 


&' 










"^ 


\ 






** 


r-' 1 




















^ 


S'~ 


















v 


^ 


























sf' 


r^. 


v^ 














>x ^ 
























^ 


t' 








V ^ 


s^ 






^ 


<* 




^^ 


\. 
















/ 


S'- 
















^^x. 


<*" 












v 


N 









0.4 
0.2 



















^ 


^s 


s 






^ 


-^ 






























,''' 










"^ 


.*"' 








^ 















FIG. 26. Effect of peak density on "directional effect." 

very small step tablet having steps each 0.025 inch long and arranged 
so as to give a series of densities simulating a sound wave of 45 cycles 
showed a definite difference of slope for the two sides of the wave as 
shown in Fig. 25. Other tablets arranged to give a similar condition 
with progressively decreasing "peak densities" show the effect to 
persist to density values as low as 0.5 peak density. (Fig. 26.) 
Microdensitometric measurements of single frequency records reveal 
a distortion of the wave shape in the manner which would be antici- 
pated from a knowledge of the direction in which they were processed. 

CONCLUSIONS 

Evidence has been presented to show that in a continuous film 
processing machine of the type used in these experiments, the uni- 



230 J. CRABTREE [j. s. M. p. E. 

directional motion of the film through the developer tends to set up a 
current within the developer solution which is parallel to the film's 
longitudinal axis, and which is sufficiently strong to dominate the 
effect of developer circulation provided by the gravity feed. 

This current results in a preferential development of any graduated 
series of light exposures such as is presented by the conventional 
sensitometer strip used in the control of sound picture development. 

This preferential development, referred to herein as a "directional 
effect," causes a distortion of the characteristic H & D curves ob- 
tained from such sensitometer exposures, unless gamma infinity is 
approached or unless means are used to obtain the necessary circula- 
tion of developer. 

The reason for this preferential development is that the reaction 
products from any image area are carried across the succeeding im- 
ages by the afore-mentioned dominant current. This trail of reaction 
products causes a retardation of the development of the images 
over which it flows. The "directional effect" is the more pronounced, 
as would be expected from a consideration of the respective time- 
gamma curves, when developing motion picture positive film in the 
customary "borax" negative developer to a relatively low gamma 
than when developing to a higher gamma in a positive developer. 
When processing in a machine in which this "directional effect" is 
known to exist, more than usual care in the exposure and processing 
of the customary type of sensitometer strip is necessary if consistent 
and representative results are to be obtained. It is recommended 
that constant exposures be given to sensitometer strips such that 
a full toe and several points on the shoulder are obtained, and that the 
strip be processed in such a way that its toe end is leading through 
the developer bath. 

Care is particularly necessary in drawing conclusions relative to the 
characteristics of the H & D curves as to the shapes of the toe or 
shoulder or the limits of the straight line portion, especially when work 
of a fundamental nature is involved. 

Where the "directional effect" is present in a film processing 
machine, it may result in the production of an asymmetric negative 
sound wave when the wave is recorded on a high gamma material 
such as positive film, and developed to a low gamma in a developer 
of the "borax" type customarily used. As a corollary to this last 
conclusion, the combination of any high gamma material and low 
gamma development will always be susceptible to irregularities in 



Feb., 1932] DIRECTIONAL EFFECTS IN PROCESSING 231 

development, and for sound recording may for the present be perhaps 
regarded only as a temporary but necessary evil. 

"Directional effect" may be eliminated by the use of any device 
that will maintain a degree of circulation which will overcome the 
current set up by the forward motion of the film itself. 

For use in machine processing, the arrangement of density areas in 
the conventional type of sensitometer strip merits consideration so as 
to provide that no density be subject to the influence of the reaction 
products of an immediately preceding density. With the present 
arrangement, distortion of the H & D curves by "directional effect" 
will be less in the case of those sensitometer strips having the larger 
physical dimensions and where the steps are as numerous as practic- 
able, thus ensuring a smaller density interval between steps. 

REFERENCES 

1 Zeit.ftir Wiss. Phot. Bd., 27, p. 236. 

2 CRABTREE, J. I., AND IVES, C. E.: "Rack Marks and Airbell Markings on 
Motion Picture Film," Trans. Soc. Mot. Pict. Eng. (1925), No. 24, p. 95. 



RESUME OF THE PROCEEDINGS OF THE DRESDEN 
INTERNATIONAL PHOTOGRAPHIC CONGRESS* 

S. E. SHEPPARD** 

The 8th International Congress of Photography was held at Dres- 
den, Germany, from August 3 to 8, 1931, inclusive. Occurring at the 
time of a financial crisis in Germany, there was a question at one time 
as to the possibility of holding the Congress at this date; but for- 
tunately, it was found possible to carry it through, and in spite of 
this unfavorable circumstance there was a very large attendance. 

The preliminary arrangements for the Congress, and the carrying 
out of these by the German committee under Professor R. Luther of 
the Technical High School of Dresden, were in the last degree praise- 
worthy and successful. 

The last day's session of the Congress was held in Berlin. After a 
visit to the magnificently equipped and sumptuously decorated new 
printing house for periodicals of the world-renowned Ullstein-haus the 
members of the Congress were taken in motorbuses to the studios of 
the Universum Film A. G. (Ufa) in Neubabelsberg. A very inter- 
esting survey was made of both the silent and sound film studios and 
laboratories. 

The work of the Congress was covered by the following sections: 

I. (a) Theoretical bases of photography. 

(6) Practice of photography. 
II. Cinematography (including the sitting of the Cine-Standards Commission). 

III. Applications of photography and cinematography in science and tech- 

nology. 

IV. History, bibliography, legal, and medical applications. 

The members of the Society of Motion Picture Engineers will be 
chiefly interested in the proceedings taking place in Sections II and 
III. However, under Section I was included the discussion of sen- 

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. Dr. Sheppard 
was the official representative of the S. M. P. E. at the Congress. 

** Eastman Kodak Co., Rochester. N. Y. 
232 



INTERNATIONAL CONGRESS 233 

sitometry, which contained several papers and reports of immediate 
interest to our Society. 

I propose to review' the proceedings of the Congress under the 
following headings: 

A. Cine-Standards Commission. 

B. Sensitometry. 

C. Miscellaneous papers of cinematographic interest. 

A. ACTIVITIES OF THE CINE-STANDARDS COMMISSION 

It should be mentioned that some time prior to the holding of the 
Congress certain changes proposed in cinematographic standards by 
the Committee of Standards of the Deutsche Kinotechnische Gesell- 
schaft had been brought to the attention of the Standards Committee 
of the Society of Motion Picture Engineers in order that it might 
express opinions on these proposals prior to the meeting of the 
Congress. I received letters and some criticisms of the German 
proposals from Mr. T. E. Shea of the Bell Telephone Laboratories, 
and a joint criticism by Messrs L. A. Jones, J. G. Jones, and E. K. 
Carver, and as far as was possible these criticisms were brought to 
the attention of the Cine-Standards Commission of the International 
Photographic Congress during their sittings and incorporated with 
the final proposals. 

The final conclusions of the Cine-Standards Commission, copy of 
which was sent to the chairman of the Standards Committee of 
this Society, and which are included in the report of the Standards 
Committee, were provisionally translated by the writer and Mr. W. 
Webb of the Eastman Kodak Company. This translation was 
based on the German text submitted to us by Dr. Erich Lehmann, 
chairman of the Committee. This statement is necessary in view of 
any possible discrepancy with the authoritative version which will be 
reproduced in the proceedings of the Congress when published, or 
with any version produced in the German technical journals. It was 
evident from the original proposals and the dimensional drawings, 
submitted as Deutsche Industrie Normen (DIN) by the Deutsche 
Kinotechnische GeseVschaft that the German proposals represented 
an attempt to bring their standard dimensions as closely as possible 
in conformity with the dimensional standards of the S. M. P. E. 
This, I think, is confirmed by the conclusions of the Committee which 
now follow: 



234 S. E. SHEPPARD [J. S. M. P. E. 

CONCLUSIONS OF THE CINE-STANDARDS COMMITTEE 

(I) Perforation Pitch. It is recommended that the length of film 
equivalent to 100 perforations be equal to 475 JI ?;o mm. This is 
the same as the standard for normal negative film. 

(II) Width of Take-up (Also Feed) Sprocket between Centers of 
Sprocket Teeth. It is recommended that the width of the take-up 
(also feed) sprocket measured from tooth center to tooth center be 



(III) Over-all Width of Take-up (and Feed) Sprocket. It is recom- 
mended that the over-all width of the take-up (and feed) sprocket be 
35.00 g;o5 mm. 

(IV) Gate Opening (Frame) for Silent Film Projectors. Recommen- 
dation is postponed. The Society of Motion Picture Engineers is 
requested to express an opinion on this question since the pamphlet 
Dimensional Standards ASA Z22 1930 contains nothing on this 
point. 

(V) Gate Opening (Frame) for Sound Film Projectors. Conclusion 
is the same as for gate-opening for silent film projectors (See No. IV 
above) . 

(VI) Tolerance in Standard Specifications. The S. M. P. E. is 
requested to amplify their standard specifications by the inclusion 
of definite tolerances in the dimensional specifications. These toler- 
ances should be expressed in metric as well as in English units and as 
far as possible should conform with the tolerances published in 
the German Industrial Standard Specifications (DIN = Deutsche 
Industrie Normen). 

(VII) Shrinkage. It is recommended that the shrinkage for 
nitrocellulose film should not exceed 1 per cent after being dried by 
suspending loosely for 240 hours at a temperature of 40 =*= 1C. in 
air having a relative humidity of 50 to 55 per cent, the air to be 
changed once or twice per hour. 

(VIII) Discrepancies in S.M.P.E. Standards. The S. M. P. E. 
is asked to express an opinion on the disagreement in the dimensions for 
the distance between teeth on sprockets as given in charts No. 6 and 
No. 7 of the pamphlet Dimensional Standards ASA Z22 1930. 

(IX) Sound Gate for Projector. The question of the length of the 
sound gate (sound slit) in projection machines must be investigated 
in all countries. There are differences between the German and the 
American standards. Special attention is called to the fact that the 
dimensions of the sound slit in projectors must be such that the slit 



Feb., 1932] INTERNATIONAL CONGRESS 235 

does not extend over the film perforations. Recommendations are 
postponed until existing differences are cleared up. 

(X) Diameter of Projection Lenses. It is recommended that a 
study of the dimensions of projection lenses be undertaken from the 
viewpoint of establishing international standard diameters. 

(XI) Definition of Safety Film. The following definition of Safety 
Film is recommended as standardized basis for all codes and regula- 
tions in all countries of the world: 

(A) Safety Film is a film which is "slow burning" and" difficultly 
inflammable." 

(B) A film can be considered to be "slow burning" if the burning 
time of a piece 30 cm. long is more than 45 seconds. In the case of 
film less than 0.08 mm. in thickness the burning time for a piece 30 
cm. long must be more than 30 seconds. 

(C) The burning time is to be determined in the following manner: 

(1) A film strip is to be used from which the emulsion has been removed by 
washing in warm water, after which the emulsion-free film base is to be dried by 
suspension in air at a temperature of 18 to 22 C. and a relative humidity of 
40 to 50 per cent for a period of 12 hours. 

(2) The test-strip has a total length of 35 cm. At a distance of 5 cm. from 
one end a mark is placed. 

(3) The test piece is suspended edge upward, if possible, in a horizontal 
position between two thin wires which are threaded through the perforations 
at intervals of not more than 32 mm. and in such a manner that the perforation 
holes utilized on the edge for this purpose do not lie opposite those so utilized 
on the other edge, i. e., the threaded holes are staggered. The wire for threading 
must have a diameter not greater than 0.5 mm. 

(4) The burning time is calculated from the time the flame reaches the mark 
5 cm. from the lighted end to the time that the whole strip has been completely 
burned. The burning test is to be carried out immediately after the film is dry 
and in a room free of draughts. The mean of at least three tests is taken as the 
final results. 

(D) A film can be considered to be "difficultly inflammable" if, 
when tested according to the method given below, it does not kindle 
(flash) at a temperature of 300C. in less than 10 minutes. 

(E) Method of making inflammability test: 

(1) The test is made in a small electrically heated furnace, the interior of 
which has the form of a vertical cylinder, with hemispherical bottom, having a 
diameter of 70 mm. and an average height of 70 mm. The opening at the top 
of the furnace is provided with a sheet iron cover in which are two symmetrically 
placed openings, one having a diameter of about 7 mm. and the other about 
15 mm. The distance between the openings is about 15 mm. from center to 



236 S. E. SHEPPARD [j. s. M. P. E. 

center. The small opening is for the introduction of an iron-constantan thermo- 
couple with a porcelain sleeve which just fits through the opening. The measure- 
ment can also be made by means of a thermometer for which stem correction 
has been made and of which the projecting stem is protected by means of a 
cork disk placed around the thermometer a short distance above the furnace 
cover. 

(2) The piece of film to be tested is hung on a U-shaped wire hook and in- 
troduced through the larger opening in the furnace cover. The solder joint of 
the thermocouple, or the bulb of the thermometer, and the center of the film 
test piece must be at the same height in the furnace which should be about 35 
mm. from the top of the furnace. 

(3) The piece of film to be tested should be 35 mm. long and 9 mm. wide and 
should have the emulsion removed by washing in warm water and drying exactly 
in the same manner as that used for preparing a piece for the burning test else- 
where described. 

(4) Before the introduction of the sample the furnace is brought to a tem- 
perature of 300 C. which must remain comparatively constant, i. e., the variation 
should not be more than = t lC. per minute. At 300 C. the sample is quickly 
introduced. 

(5) Before repeating each test the cover of the furnace must be removed and 
the products of combustion completely removed by means of an air blast. 

(XII) Edge Marking of Safety Film. It is recommended that, 
upon safety film, having a width of more than 34 mm., there be 
placed a special characterizing mark which will be visible and recog- 
nizable when the film is spooled in the form of a roll. As a means of 
accomplishing this it is recommended that the edge of the film be 
provided with a thin protective coating which hinders alteration in 
the emulsion layer during the subsequent processing operations. 

On those matters for which no conclusion was arrived at, such as 
No. (IV), the gate opening for silent film projectors, No. (V), for 
sound film projectors, it will be noticed that an expression of opinion 
is requested from the S. M. P. E. It should also be noticed in regard 
to No. (VI) that the Society is requested to change its previous 
policy by including definite tolerances in their dimensional specifica- 
tions. 

The chief difference on any one point from the Society's definitions 
is in regard to the definition of safety film. In the writer's opinion a 
burning test based on the horizontal burning is more reliable than one 
based on the test piece in the vertical position. In connection with 
these definitions it should be recalled that Dr. Lehmann's proposals 
were made in connection with a meeting of the industries interested 
on the one side and the governing bodies in Germany on the other 



Feb., 1932] INTERNATIONAL CONGRESS 237 

side, in connection with the safety regulations for cinema shows, and 
in particular a draft act known as a Narrow Film Act. It may be of 
interest to quote its first three paragraphs: 

NARROW FILM ACT 
Paragraph 1 

Narrow films, in the sense understood by this act, are film ribbons which are 
intended for taking pictures, writings and the like, and whose width is less than 
34 mm. 
Paragraph 2 

Narrow films must not be easily inflammable nor easily combustible. Easily 
inflammable and easily combustible narrow films must not be manufactured 
at home or be introduced from abroad. 
Paragraph 3 

Easily inflammable and easily combustible narrow films must not be brought 
into the market nor introduced in the trade after the coming into force of this 
act. Also, their application in cinema theaters, public buildings, halls, or picture 
palaces is prohibited. 

It will be seen that the definition given in regard to the specifica- 
tions of safety film refer particularly to paragraphs 2 and 3 by way of 
actual definition and interpretation of the terms "not easily inflam- 
mable" and "not easily combustible," or alternatively, the terms 
"slow burning" and "difficultly inflammable." 

The conclusions reached by the Committee are necessarily held for 
six months for approval by the national committees for the Inter- 
national Congress of Photography. 

B. SENSITOMETRY 

In regard to sensitometric standardization, several important 
developments occurred. First, the other national committees on 
sensitometric standardization accepted the light source and filter 
proposed by the American Committee at Paris, 1925, and accepted by 
the British in 1928. In the meantime, no definite agreement had been 
reached, nor indeed had very definite proposals been made on the 
subjects of sensitometers or exposure meters, development, density 
measurement, and methods of expressing sensitometric results, al- 
though much discussion and controversy on this subject had taken 
place. At the present Congress, a body of recommendations for sen- 
sitometric standards was put forward by the Deutschen Normen- 
ausschusses fur Phototechnik, which endeavored to cover the latter 
questions and bring the subject of sensitometric standardization 
into the industrial field. It was stated by the German committee 



238 S. E. SHEPPARD [J. S. M. P. E. 

that this action had been forced on them by difficulties arising from 
indiscriminate and uncontrolled placing of speed numbers on photo- 
graphic sensitive goods, a situation which was summarized at the 
Congress by the term "Schemer-inflation." 

The gist of these recommendations was as follows: 

(a) Acceptance of the light source and daylight filter as proposed by the 
American commission. 

(6) As exposure meter, a density step-wedge combined with a drop shutter 
accurate to Vo second. 

(c) Brush development in a tray with a prescribed solution of metol-hydro- 
quinone according to a so-called "optimal" development. 

(d) Expression of the sensitivity by that illumination at which a density of 
0.1 in excess of fog is reached. 

(e) Density measurement shall be carried out in diffused light according to 
details to be discussed later. 

These proposals aroused a very lively discussion. The American 
and the British delegations criticized the proposals both as a whole 
and in detail. As a whole they considered that the time was not 
ripe for application of sensitometric standards to industrial usage. 
In matters of detail they criticized the proposed employment of a step- 
wedge, and the particular sensitivity number proposed. The latter 
approaches very roughly the idea of an exposure for minimum gradi- 
ent, but even such a number is not adequate for certain photographic 
uses of certain materials. 

The upshot of the discussion was that the German proposals in 
somewhat modified form are to be submitted simply as proposals of 
the German committee for sensitometric standardization to the 
various national committees for definite expression of opinion within 
six months of the expiration of the Congress. Further, in case of 
general approval of these recommendations by the other national 
committees, that a small International Committee on Sensitometric 
Standardization shall, within a further period of six months, work out 
a body of sensitometric practices for commercial usage. 

In this connection it should be noted that it was agreed that both 
the lamps and filters and exposure meters should be certified as within 
certain tolerances by the national testing laboratories of the countries 
in question. 

BRIEF REVIEW OF PAPERS PRESENTED 

It is obviously impossible, as it would be undesirable, to review 
in extenso the papers of cinematographic interest presented at the 



Feb., 1932] INTERNATIONAL CONGRESS 239 

Congress. These papers will be published in full in the Proceedings 
of the 8th International Congress of Photography, and it is for the 
benefit of those who may wish t6 study them more fully there that I 
am giving the following references. 

The following papers represent those of general interest to cine- 
matography on its technical side, although not necessarily cine- 
matographic: 

A paper by W. Dziobek, of the Physikalisch-Technische Reichsanstalt, dealt 
with the use of the tungsten vacuum lamp in sensitometric measurements. It 
points out that for this purpose the following data should be known: 

(1) The amperage at which the radiation has a color temperature of 2360 K. 

(2) Light intensity in international candles at the amperage determined by 1. 
It was concluded that if the color temperature can be reproduced to within 

10 the resulting error in actinic intensity amounts to only 0.5 per cent. Curves 
were given showing alteration of color temperature of a series of tungsten vacuum 
lamps over a lengthy period of burning. The constancy was found sufficient 
after a period of running of from 80 to 100 hours. If run for at least 80 hours 
at a normal load only exceptionally should a falling-off of 1 per cent occur for 
100 hours' further running. 

Color cinematography was considered in only two communications, and these 
both in the nature of semi-popular, general lectures. Professor J. Eggert of 
Leipzig gave a special lecture on the present position of color cinematography 
illustrated by examples covering two- and three-color additive processes, two- 
color subtractive processes, direct and indirect screen processes. 

Mr. Thorne-Baker gave a paper illustrated by examples of the Spicer-Dufay 
process of color cinematography. This consists in preparing upon a continuous 
band of film base a three-color matrix or "screen" having 900 or more colored 
rectangular areas per square millimeter. The "screen" is then coated with 
emulsion and exposure is made through the support and "screen." It may be 
developed as a negative or as a reversed positive. Methods of making copies 
at the standard rate of 800 pictures per minute were described. 

In connection with sound film and sound pictures papers of both general and 
special interest were presented. 

Dr. E. Goldberg, of Zeiss-Ikon, Leipzig, gave an extremely well demonstrated 
and illustrated popular lecture on "Fundamentals of the Talking Films." 

O. Sandvik and L. A. Jones of the Eastman Kodak Company presented a 
review of the talking film. 

A paper by H. Thirring dealt with "Sound Reproduction by the Selenophon 
Process" which has been developed by the Austrian Sound Film Company. 
The modulation of the light beam is effected by a string oscillograph in which 
a metallized thread stretched in the field of an electromagnet cuts the real image 
of a luminous slit at a small angle and in a position of rest covers half of it. The 
telephonic currents from the microphone of the taking studio, after suitable 
amplification, are conducted through the thread which is thereby set in oscillation 
and modulates the length of the free part of the line of light. By the registration 
of this line of light of variable length on a film moved perpendicularly to the 



240 S. E. SHEPPARD [J. s. M. P. E. 

length of the line there results a variable width sound record. It is stated that 
the process has been adapted to the reproduction of a photo-phonic gramophone, 
the phonograms consisting simply of paper, copies of sound film records. 

A paper by C. R. Keith, of the Western Electric Company of London, dealt 
with "Distortion Factors in Sound Reproduction by the Intensity Process." 
It was pointed out that the differences which exist between the actual gamma 
value of a sound record and the gamma of a sensitometer strip developed at the 
same time could be traced to (1) the effect of the different light intensity and 
reciprocity failure; (2) the Callier effect; and (3) the color effect resulting 
from incomplete correspondence of the spectral composition of the light sources 
used. The author described methods for overcoming these difficulties. 

R. Thun dealt with "Technical Problems of the After-Synchronization of 
Films" (Dubbing). He analyzed the problem as follows: 

(1) Determining the desired association of sound and picture. 

(2) Approaching the sound sequence as closely as possible to the fixed scheme. 

(3) Detection of residual defects. 

(4) Removal of these by correction of the picture or the sound sequence. 

It is claimed that better results are obtained by corrections applied to the 
picture rather than to the sound records. 

R. Schmidt, of the Agfa Company, discussed "Ultra-Short-Exposure Sensitom- 
etry and Reciprocity Failure in Special Relation to the Making of Sound Films 
by the Method of Variable Exposure Time." For periods of illumination from 
Vioo to 1 / 8 o,ooo second it was observed on decreasing time of exposure that a flatten- 
ing of the gamma value of the characteristic curve occurred. It is of special 
importance for the intensity process with variable time of exposure to know 
the actual gradation curve of the taking exposure in order to compensate for 
distortions. The author has applied the form of representation (formerly given 
by Arens and Eggert) of the relation of density to light intensity and time of 
exposure by means of density-intensity-time surfaces. 

New results in x-ray cinematography were described by K. Jacobsohn, scien- 
tific editor of "Photographische Industrie." This dealt particularly with ex- 
periments made with Dr. V. Gottheimer of the Pankow Hospital, Berlin. They 
were made by the indirect method, namely, cinematographing the image on a 
fluorescent screen. The improvements discussed consist in: 

(1) Taking camera having a special claw mechanism by which the film is 
kept longer at a standstill at the gate at the expense of the time of exposure. 

(2) A lens of great aperture //1. 25 consisting of two pairs of cemented glasses. 

(3) Special fluorescent screen, resembling an intensifying screen. 

(4) Ultra-sensitive film. 

The value of x-ray cinematography as compared with subjective observation 
of movements of internal organs was discussed. 

Of papers of more specialized character the following may be mentioned: 
"A Micro-Cinematographic Outfit" described by H. Linke, constructed by the 
Askania-Werke of Berlin-Friedenau, and which was on view at the exhibition 
associated with the Congress. 

F. Beck of the same firm described "Cinematographic and Photographic Meth- 
ods for Investigating Rapidly Recurring Processes." The operations of high- 



Feb., 1932] INTERNATIONAL CONGRESS 241 

speed cinematography were described in detail, as well as the use of rotating 
cameras and series cameras. The application of these methods to the study of 
explosions, operation of explosion motors, combustion processes, spark phe- 
nomena, explosive tests, and lightning were discussed. 

Another paper on somewhat the same subject was by W. Ende, of the A. E. G., 
Berlin, entitled "New Results in the Application of High-speed Cinematography 
to Technical Research." This discussed the special requirements in regard to 
speed and registration in the design of high-speed cine cameras for technical and 
scientific research. It was considered that the Thun "Zeitdehner" (time 
stretcher) was the best instrument for taking a large series of pictures on a running 
band of film. Various modifications and accessories of the Thun "Zeitdehner" 
were described, such as apparatus for regulating the speed of taking on the film, 
an optical indicator, and an automatic release by the camera. A method of in- 
creasing the speed of taking was described which allows the number of pictures 
to be increased from 6000 to 30,000 per second. The paper was illustrated by 
a film showing the high-speed study of mechanical movements, of arcs and spark 
phenomena, with exposures ranging from 1000 to 30,000 per second. 

9TH INTERNATIONAL CONGRESS OF PHOTOGRAPHY 

At the concluding business meeting of the Congress, the writer, 
in the names of the Society of Motion Picture Engineers and the 
Optical Society of America, offered a provisional invitation to the 
Congress to make its next meeting (1934) in North America. This 
proposal was received with much appreciation, but with definitely 
expressed doubts as to its feasibility. It is hardly to be denied 
that a meeting on this side is desirable. Eight of these Congresses 
have now taken place in Europe. The last three post bellum Con- 
gresses were held in Paris, London, and Dresden. They have ex- 
emplified in their own field the unity of science in western culture, 
in the face of national and linguistic differences. If the International 
Congress of Photography is to be truly international, and not merely 
European, it is essential that it should meet before long on this side 
of the Atlantic. Our technical societies, directly or indirectly con- 
cerned with photography, and the great American industries of cine- 
matography and photography, will assuredly honor themselves and 
materially assist photographic advance by helping to bring about an 
American meeting of the International Congress. I call to your at- 
tention that this would be the first meeting of the Congress under its 
new name, since at the conclusion of the Congress it was decided to 
change the name of the Congress from the International Congress of 
Photography to International Congress of Scientific Photography and 
Cinematography. It is my sincere hope that this Society will do all 
in its power to make the invitation effective. 



COMMITTEE ACTIVITIES 

REPORT OF THE PROJECTION SCREENS COMMITTEE* 

The first report of the Projection Screens Committee was published 
in the September issue of the JOURNAL. It was to have been read 
at the May Convention of the Society but unfortunately the copies 
shipped to Hollywood by air mail were lost in transit. It dealt 
with, in some detail, the manufacture, installation, and maintenance 
of screens, and their light-reflecting and sound-transmitting proper- 
ties. Curves were given to illustrate the reflection characteristics 
for the three types: diffusing, metallic, and beaded. Sound re- 
quirements and test methods were discussed at length. 

It is, of course, our hope to consider screens from every possible 
angle of interest to the Society. At the present time we are able 
to report further progress on the program we originally formulated. 
We have some data on deterioration of screen surfaces, enough to 
indicate that a serious condition exists. The troublesome problem 
of determining the optimum illumination for screens has been given 
considerable attention, and some interesting information on rear 
projection screens, and incidentally rear projection, has been ac- 
cumulated. This material follows. 

LIGHT REFLECTION 

That screens lose their reflective power with use is common knowl- 
edge. However, reliable data as to the magnitude of this loss have 
never been accumulated. We have made a beginning in this direc- 
tion. The few results we have had the time to obtain indicate the 
range of variation and the really serious extent of the deterioration. 

Our measurements were made with equipment constructed by 
one of the members of this Committee. The apparatus consisted 
of a metal tube 4 inches in diameter, holding a lamp operating at a 
color temperature of 2360 K. Concentric with this first tube, 
there was a second narrower one with a viewing aperture at one 

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. 
242 



PROJECTION SCREENS COMMITTEE 243 

end, in which it was possible to insert the photometer unit of a Mac- 
beth illuminometer. The light source was an automobile lamp 
placed so that the angle of light incident upon the screen was approxi- 
mately 3 degrees with the normal, which was the viewing angle. 
Between the viewing aperture and the screen was placed a blue 
filter such as to make the color of the light entering the photometer 
correspond to 5000 K. The light in the Macbeth comparison 
lamp was corrected with a similar filter in order to eliminate color 
difference in making the photometric balance. A battery of five 
dry cells made the apparatus entirely portable and independent of 
external power. 

Measurements were made hi several convenient theaters. The 
device was placed against the screen which was observed through 
the photometer inserted in the aperture. Obviously, the data are 
restricted to only one angle. It is felt that the loss of reflection 
which occurs at one angle will indicate approximately what occurs 
at other angles. The results are summarized in the following table. 
The original values for these screens ranged from 77 to 85 per cent. 

Reflection at 3 Degrees from Normal 

Sample Per Cent 

A . Broadway Theater 

(In use 18 months) . 45 

B. Auditorium; New York, N. Y. 

(In use occasionally for 3 years) 48 

C. Broadway Theater 

(In use 18 months; lately reprocessed) 70-80 

D. Broadway Theater 

(In use 9 months; reprocessed 3 months) 64 

E. Hoboken Theater 

(In use occasionally 9 months) 80 

F. Review Room 

(In use 9 months) 76 

It will be noticed that deterioration is not very consistent. How- 
ever, we should expect it to vary widely, depending on the conditions 
surrounding the use of screens. The valuable results obtainable 
from surface reprocessing are demonstrated by case C. In addi- 
tion to a possible degradation of picture quality there will be a finan- 
cial loss accompanying deterioration of reflecting ability. Some 
idea of the possibilities may be grasped from the following table. 



244 PROJECTION SCREENS COMMITTEE [j. s. M. p. E. 

The figures are based on a hypothetical decrease of 20 per cent 
in reflection, a serious loss. It is assumed as a first approximation 
that this corresponds to a 20 per cent waste of electric power. Other 
assumptions are: operation, nine hours daily, and energy cost, five 
cents per kilowatt hour. These are common conditions. 



Type of Lamp 


Amperage 


Weekly 
Current 
Cost 


Weekly 
Loss 


Low intensity arc 
Hi-Lo intensity arc 
High intensity arc 


25 
75 
120 


$ 8.75 
26.00 
42.00 


$1.75 
5.20 
8.40 



Obviously, if projection occurs only a few hours daily or weekly, 
the loss is not serious. However, it is not difficult to imagine a case 
where replacement of a screen would soon pay for itself by the sav- 
ing of power required for illumination. One example is the theater 
given as case A above, which runs approximately thirteen hours 
every day. Another case when difficulty arises occurs when the 
projection outfit is operating at the limit of its capacity and is un- 
able to supply sufficient light to overcome the loss of reflective 
ability of the screen. 

SCREEN ILLUMINATION 

Upon the appointment of the .Projection Screens Committee, 
President Crabtree stressed the need for recommendations on the 
amount of screen illumination required for motion picture presen- 
tations. This complex subject has received a great deal of atten- 
tion in the past, being one of the oldest of projection problems. Much 
scattered work has been done without the achievement of standardiza- 
tion or a complete realization of the factors involved. Among 
these factors are included the simplification of studio lighting and 
printing control as well as projection illumination technic. We 
will review hastily a few of the facts that are known, before describing 
a series of tests conducted at the meeting of the New York Section, 
Friday, September 25, 1931. 

Factors that must be considered include visual acuity; flicker; 
other physiological factors; fidelity of brightness, contrast, and 
tone reproduction; and auditorium lighting. A complete unravel- 
ing of these is impossible but we may analyze them to some extent 
to obtain a better understanding of the problem. 



Feb., 1932] PROJECTION SCREENS COMMITTEE 245 

A picture projected on a plane surface will be seen to consist 
of a grouping of areas of different brightnesses, that is, it is merely 
a pattern of contrasts. The relative brightness of the images must 
be presented much as the subjects are in actuality. For our pur- 
poses it is not necessary to discuss so much the relations involved 
among the areas as it is the intensity of light with which the picture 
as a whole is to be projected, that is, the absolute values of bright- 
ness. We seek to learn how brilliant bright objects must be, how 
dull the dark subjects may be and yet be discernible. 

Obviously it is not possible to reproduce on the screen values of 
brightness as they occur hi nature. To a large extent this is not 
necessary and often not even desirable. One purpose of a motion 
picture is to create artificially an impression which will be accepted 
as a satisfactory illusion of reality. More than that, it aims to con- 
vey a story, using its own devices. The brightness element, together 
with size, depth, and color, is secondary, being subordinate to the 
story and continuity. Hence, it is not necessary that sky scenes be 
shown with clouds as lustrous as clouds are, human faces as bright 
as they are in every-day life, deep shadows as profound as they often 
are. What is essential is not so much faithfulness to actuality as it 
is adaptation of illumination to achieve a smooth vivid portrayal 
of the story. This is fortunate in as much as we possess no light 
sources capable of producing on a screen brilliancies comparable 
with those under direct sunlight. Nevertheless there may be some 
instinctive demand for reasonable fidelity in brightness reproduc- 
tion. 

The pictures on the screen should be easy to see under conditions 
of illumination existing in theaters. The projected image should 
be the brightest area in the theater to facilitate concentration. 
However, in addition to this psychological element, there is another 
practical requirement. The auditorium should be provided with 
as much light as is consistent with preserving satisfactory detail in 
the picture, and the intensity should increase with the distance 
from the screen. This light should be sufficient to mitigate screen 
glare and permit easy finding of seats. There should be no sudden 
change at any point, as sharp contrasts are harmful to the eye. Stray 
light falling on the screen must be kept to a minimum in order to 
preserve picture contrast. Clearly, if the stray light should equal 
the illumination on the screen corresponding to a shadow, the shadow 
would disappear. 



246 PROJECTION SCREENS COMMITTEE [J. S. M. p. E. 

The lower limit of screen brightness should therefore be deter- 
mined by the light reaching the screen from the auditorium. There 
is no criterion for the maximum desirable amount of illumination 
corresponding to the highlights. We do know, however, that with 
the auditorium in a darkened condition it would not do to have 
too bright a screen as this would be physiologically harmful. 

Desirable screen brightness is dependent on all these variables. 
Only by analysis of judgments drawn from many observers subject 
to varied, controlled conditions will it be possible to determine the 
optimum relations. 

In an endeavor to obtain more information on this subject, we 
conducted our tests at the meeting of the New York Section. This 
meeting afforded an excellent opportunity in as much as there was 
present a body of trained men who would readily understand our 
aims. We did not expect conclusive results from our tests, but 
regard them as a preliminary step in the investigation. Obviously, 
a complete study of all the factors would require the time of many 
men over a period of months. 

In these tests we used two projectors, one a hi-lo and one a low 
intensity arc, setting these to produce known values of screen illumi- 
nation. Two types of arc were employed to determine whether 
different color characteristics affect the amount of light judged de- 
sirable. There was no illumination in the auditorium other than 
that supplied by screen reflection. It would have been interesting 
to vary the lighting also, but the time at our disposal necessitated 
restriction of the variables. Two reels of film were used, one with 
a large percentage of brilliant scenes in it, such as outdoor shots, 
the other consisting of interiors, emphasizing human features and 
shadows. We wished to learn whether different amounts of light 
would be found desirable for different types of subject-matter. 

The arc light intensity was varied by means of wire filters inserted 
in the projection machine behind the condenser lens. Four settings 
were used. The first setting was 68 per cent of the maximum, 
which was the second setting. The third and fourth settings were 
50 and 25 per cent, respectively. The low intensity machine was 
first used for both reels and was followed by the hi-lo intensity arc. 
There were present 61 observers, most of whom commented on the 
projection on questionnaires which were distributed among them. 
Their findings are summarized in the following table. The bright- 
ness values are without film and with the shutter running. 



Feb., 1932] 



PROJECTION SCREENS COMMITTEE 



247 



Reel 1 
Interiors 



Reel 2 
Exteriors 



Low Intensity 



Brightness 


4.7 


Foot 

7 


Lamberts 
3.5 


1.7 


4.7 


Foot 

7 


Lamberts 
3.5 


1.7 


Glaring 





3 








10 











Bright 


2 


8 








20 


4 








Preferred 


9 


17 





1 


12 


18 








Acceptable 


26 


25 


3 


I 


16 


29 


7 





Dull 


18 


4 


41 


12 





8 


39 


15 


Dark 


2 





14 


44 








12 


43 



Hi-Lo 





11.5 


Foot 
17 


Lamberts 
8.5 


4.2 


11.5 


Foot Lamberts 
17 8.5 


4.2 


Glaring 


10 


20 








13 


6 








Bright 


14 


14 


1 





20 


10 


2 





Preferred 


9 


12 


11 


1 


7 


22 


7 





Acceptable 


14 


10 


18 


7 


15 


16 


8 


3 


Dull 


9 


1 


19 


24 





4 


33 


20 


Dark 


1 


1 


8 


24 








8 


35 



Screen reflection factor: 80 per cent. 

Screen size: 9 by 12 feet. 

Distance from screen: from 27 to 55 feet 

Viewing angle* 90 =*= 30 degrees with screen. 

Auditorium illumination: 0.02-0.5 foot candle. 

Brightness of screen surroundings: 0.1-0.9 foot lambert. 

19 Observers expressed a preference for the color of the low intensity lamp; 

17 preferred the hi-lo. 
A foot lambert is the brightness of a perfectly diffusing surface illuminated 

by one foot candle. 

Analysis of Results. Under the circumstances we cannot be too 
positive in our conclusions from these tests. It will be sufficient to 
point out tendencies and possibilities. To obtain decisive results 
it would be necessary to perform repeated and varied experiments 
lasting over a period of time. Admitting the limitations, we may 
proceed to interpret the data. 

With reel 1 and the low intensity lamp the reactions were just 
what might be expected. A brightness of 7 foot lamberts was found 
to be quite acceptable. This reel consisted of views of a string 
orchestra, the players being dressed in dark, formal clothes. The 
brightness on the screen was of the same order of magnitude as those 
existing at an actual performance of such an orchestra. Obviously, 



248 PROJECTION SCREENS COMMITTEE [J. S. M. P. E. 

we do not know that this value would have been preferred to a higher 
one, which our facilities did not permit. 

The results obtained with this reel and the high intensity lamp are 
in fair agreement with those for the low intensity. A brightness 
of 17 foot lamberts is too great for such an indoor scene projected 
in a darkened auditorium. A value between 8.5 and 11.5 is indi- 
cated as perhaps the most acceptable. 

Reel 2 consisted of comparatively brilliant outdoor scenes. It 
was shown after the reel of indoor scenes and it is supposed that the 
first reaction of the audience was to pronounce the illumination 
bright. After sufficient time had elapsed for ocular accommodation, 
a greater brightness was found acceptable and, in the case of the high 
intensity lamp, preferred. The light intensities on the screen were 
naturally far below those at which the original scenes would have 
been viewed. 

One conclusion is that it is necessary to vary the light intensity 
for different types of prints, although it is theoretically possible to 
select one light intensity and maintain it by recording scenes on a 
sliding photographic scale, each value of brightness to have a definite 
constant position on this scale. The optimum value of brightness 
according to these tests should be a compromise between the ex- 
tremes of 7 and 17 foot lamberts, the mean of which is 12. This 
is somewhat higher brightness than is customary. 

REAR PROJECTION 

Historical. Rear projection is not new; it has been used for 
fifteen years in Germany, France, and England. In this country 
we are all familiar with the small projectors used in public places 
for advertising, demonstration, and stock quotations. Application 
to the theater was delayed by two difficulties: one, the lack of a 
suitable translucent material, and the other, of an efficient distor- 
tionless wide angle lens. Within the past six months several small 
theaters have opened in New York to show newsreels and sliort 
subjects on a rear projection screen. 

Mechanics. There are several possible materials for use as rear 
projection screens. The more common are dental rubber, treated 
silk, ground glass, celluloid, and a gelatin composition. The last 
is one which is being used on a large scale. Glass screens have 
a satisfactory transmission characteristic but the large sizes are 
heavy and difficult to protect. Celluloid screens would be satisfac- 



Feb., 1932] PROJECTION SCREENS COMMITTEE 249 

tory if it were not for their fire hazard. All rear projection screens 
have the disadvantage that large uniform areas of material must 
be used. They differ from front projection screens in this respect, 
for the latter are sewed together from strips of standard width. 
The process of manufacture of the gelatin screen is as follows: 

On a heated table is poured a hot gelatin solution, over which is 
stretched smoothly a fine silk fabric which is pressed into the gela- 
tin. The combination is allowed to cool slowly about twenty -four 
hours, and is then placed on a rack to dry for seventy- two hours. 
Care must be taken to keep water from touching the screens as the 
composition is soluble in water. The screens may be cleaned with 
alcohol. They can be furnished in any desired color but at present 
a slight bluish tint is standard. 

Installation. It may be of interest to point out several facts about 
the installation of rear projection apparatus as it is done in the new 
small theaters. Standard apparatus is used, two machines being 
mounted about 8 feet behind the screen at an angle of 45 degrees 
with each other and 22 Vz degrees with the screen normal. Each 
lens is approximately 7 inches off the screen axis. 

The width of screen that is possible is determined by its distance 
from the projection lens. The rule is that 1 foot of width is possible 
for every foot of separation between the screen and the 1-inch focal 
length lens that is employed, 8 feet in this case. 

There is a general impression that film as projected over these 
machines must be reversed. This is not so, as a prism is employed 
to reverse the image on the screen and to bend the light rays through 
an angle of 22 l / 2 degrees. The prism is mounted immediately 
ahead of the negative projection lens. 

The screen is mounted about 5 inches above the head of an observer 
in the first row. This makes possible the installation of a horn or 
baffle loud speaker beneath and on a line with the screen. It must 
be pointed out that this position for the speaker is not quite correct 
for furnishing the proper illusion, which, however, is yet acceptable 
in the front rows to the ordinary observer. At the rear of the theater 
the effect is quite good, in as much as the auditorium is small and 
sound mixing helps create the correct impression. 

One advantage of the rear projection installation may be pointed 
out. It requires less vertical space and no specially dimensioned 
auditorium. Hence it is possible to employ as theaters enclosures 
similar to small stores. 



250 PROJECTION SCREENS COMMITTEE [j. s. M. p. E. 

Light Transmission. The light transmission may be varied to 
meet different requirements. We have already seen that trans- 
mission may be made to favor any particular color. It also may 
be made to give several different types of distribution. By proper 
processing, the distribution is made more uniform, and hence satis- 
factory for viewing at wider angles. It must be expected that there 
will be an additional loss of contrast as compared with front pro- 
jection because of the introduction of another translucent surface, 
which adds to the flare effect. 

Illumination.- Since the screen is light transmitting, the light 
intensity in the auditorium can be considerably higher than in the 
ordinary theater during a performance. It has been stated that the 
auditorium is illuminated to about 30 per cent of average theater 
full lighting. Nevertheless, it is necessary to take precaution to 
keep light from falling on the screen, in as much as there is some 
slight reflection from the surfaces. High auditorium illumination 
means that confusion in seating is practically eliminated. For types 
of theaters where patrons are continually passing in and out, it is 
very desirable to have considerable light. However, it must be re- 
membered that a partially lighted auditorium tends to prevent 
patrons from "living" through a feature presentation, since it makes 
one too conscious of his immediate surroundings. In a theater 
showing newsreels and short subjects, this is not objectionable. 

For much of the above information on rear projection we are 
indebted to Mr. W. Mayer and the Trans-Lux Movies Corporation. 

S. K. WOLF, Chairman 
D. S. DE'AMICIS W. F. LITTLE 

F. M. FALGE A. L. RAVEN 

H. GRIFFIN C. TUTTLE 

DISCUSSION 

PRESIDENT CRABTREE : The work of this Committee points the way in which 
a committee can do real research work. They did not have to have a research 
laboratory in which to make these tests. They used the available research labora- 
tory, which was the membership of the Society. I congratulate the Committee 
on this pioneering effort in cooperative research. 

It is very interesting to find that it seems to be necessary to have a greater 
screen brightness for the outdoor shots than the indoor ones. On second thought, 
it is reasonable. Probably the matter could be taken care of by giving a uniform 
flash exposure to the interior scenes, or they could be printed a little heavy. 

With reference to the brightness test I should like to point out that the figures 
show approximately ten foot candles as the minimum desirable brightness of a 
picture. In previous years we have made numerous tests of screens, and find that 



Feb., 1932] PROJECTION SCREENS COMMITTEE 251 

the average lies between three and five foot candles; and with low intensity light 
sources, three foot candles. In the studio laboratories where prints are analyzed 
by the studio personnel, intensities of about thirty foot candles are used ten 
times the intensities used in the theaters. This is the cause of dark prints and the 
troubles that go with them, such as overloading. Dirty screens also require over- 
loading, causing additional loss of picture quality. 

MR. FARNHAM: In connection with the data on reflection factors of screens, 
the figure of eighty per cent appears to be very high. I should like to ask if that 
is absolute reflection, i. e., incident light to total reflected light or is it the ratio 
of reflected light after a period of use to that of a new screen? 

MR. WOLF: The measurements of reflection factor were made as soon as the 
process was completed; each time a comparison was made with the standard. 

MR. FALGE: It happens to be the brightness at the normal which is hi ques- 
tion rather than the total reflection value. 

MR. FARNHAM : That is an extraordinarily high value, and that is why I asked. 

PRESIDENT CRABTREE: Have the experiments of the Committee, Mr. Wolf, 
gone far enough that we can begin to think of recommending a standard of screen 
brightness? 

MR. WOLF: The data collected at the demonstration at Bell Laboratories 
proves more or less conclusively that there are certain limits to be considered. 
We cannot say definitely what they are, but they probably lie between seven and 
thirteen foot candles. We definitely believe that any picture having a brightness 
less than seven foot candles is certainly too dull ; and any picture having a bright- 
ness greater than thirteen foot candles is glaring and disagreeable to look at. 

PRESIDENT CRABTREE : Can any one explain why a value of thirty is used in 
the screening room? 

MR. FALGE: In an article published by Mr. Huse some time ago, describing 
some tests of Hollywood screening rooms, he gave the value as thirty foot candles. 
It is expected that the intensity will always be high in the screening rooms, unless 
deliberate attempts are made to keep it within reason, because the picture is 
always small and the light intensity is greater than in the average theater. 

PRESIDENT CRABTREE: Were Mr. Huse's measurements strictly comparable 
with yours? In other words, has the Committee first of all found a method 
of getting an absolute measure of this reflection value? Does your figure of ten 
correspond with a similar figure in Hollywood? 

MR. FALGE: I think you will find considerable variation among the figures 
that have been collected; but I think it is sufficiently important, even with an 
error as great as twenty-five per cent, to show that the values in the theater differ 
considerably from those in the studios. 

MR. FARNHAM: As a result of some work that I did a number of years ago on 
screen brightness, I found that there is also a relation between the picture size and 
the screen brightness. Smaller screens should be brighter for the same projected 
picture, so that whatever intensities we recommend for the studio viewing rooms, 
they must be corrected for the size of the picture. 

PRESIDENT CRABTREE: That was also observed when we were making wide- 
film experiments. We did not need as great a brightness as with the smaller pic- 
tures. 

MR. GAGE: May I ask if the foot candles are measured with a machine sta- 



252 PROGRESS COMMITTEE WORK [J. s. M. P. E. 

tionary with the shutter open, or with the shutter running and with no film? 

MR. WOLF: We have data under all conditions. In preparing the tests we 
made measurements both with the shutter standing still and with it in operation. 
We made the measurements, also, of auditorium illumination and other quan- 
tities. The screen reflection factor was eighty per cent, and the size of screen 
was nine by twelve. The auditorium illumination varied from 0.2 to 0.5 foot 
candle. The amperage of the high intensity arc varied from 7 to 4.2; of the 
lower intensity, from 7 to 4.7. 

PRESIDENT CRABTREE: What were the limits of variation due to screen size? 
Do you recall, Mr. Farnham? 

MR. FARNHAM: The smallest screen used was approximately four feet and 
the largest twenty-two feet, a linear ratio of one to five and one-half. However, 
the ratio of brightnesses was more nearly two or three to one, the smaller picture 
requiring the higher intensity, but it was by no means an inverse ratio. 

PRESIDENT CRABTREE: Suppose a value of seven were required for a twenty- 
foot screen, what would be the value for a four-foot screen? Would it be greater 
than thirteen? 

MR. FARNHAM: As near as I can recall, the smaller picture would require 
two to three tunes the intensity ratio. 

PRESIDENT CRABTREE: Is the Committee considering the effect of screen size? 

MR. WOLF: Yes, it is; but sufficient time was not available. 

PRESIDENT CRABTREE: If any of you are in New York I would recommend 
that you visit one of the Trans Lux theaters where pictures are projected from the 
rear of the screen. The most amazing thing is that the brightness level in the 
theater is as high as it is in this room, and yet the picture is adequately bright. 

MR. GAGE: With a small screen close by or a large screen far off, both sub- 
tending the same angle to the eye, and with the same foot candles of illumination 
would not this give equally desirable results on both screens? If so, it is necessary 
to relate the distance of the observer to the screen size rather than simply say 
that a twenty-foot screen requires so many foot candles, and a thirty-foot screen 
so many foot candles, etc. 

PRESIDENT CRABTREE : That would depend on the opacity of the atmosphere. 

MR. WOLF: We did find a difference in the reactions of viewers as they moved 
away from the screen. But the brightness is the same whatever the distance may 
be. 

PRESIDENT CRABTREE : Not if there is absorption, and the air is full of smoke. 

MR. WOLF: That effect is not appreciable. 

PRESIDENT CRABTREE : I urge the Committee to push forward the experiments 
as rapidly as possible, because I am anxious that our Society should be the first 
to propose a definite standard of screen brightness with the necessary qualifica- 
tions due to the various factors involved. 

ORGANIZATION OF PROGRESS COMMITTEE WORK 

For three years the past-chairman of the Committee has assisted in 
the preparation of the semi-annual report, and it has occurred to him 
that a re'sume' of the program of organization may be of some value 



Feb., 1932] PROGRESS COMMITTEE WORK 253 

to future chairmen. The following notes represent a description of 
the plan of organization of the work of the Committee. 

Membership of the Committee. It is very important in selecting 
members of the Committee to choose men who are representative of 
various departments of the industry. Such phases of the industry 
should include: film manufacture, lens design, camera work, and 
sound recording technic, studio illumination, laboratory processing, 
sound reproduction, theater construction and operation, and applied 
cinematography. Besides representatives in the United States, men 
should be selected from each country or part of the world where a 
well-developed motion picture industry exists, as well as where re- 
search on cinematographic problems is in progress. 

The widely separated geographical position of the members of the 
Committee makes it unfeasible to hold meetings so that all the com- 
mittee work must be handled by correspondence. Each member 
should be instructed carefully relative to the scope of the field which 
he is to cover in his semi-annual report to the chairman. It is very 
desirable to distribute the abstracting work of the Committee mem- 
bers, and separate journals which are pertinent to the nature of their 
own work should be assigned to each member. 

The reports from Committee members may be composed of any one 
of the following types of information: 

(1) Abstracts of journals. 

(2) Personal appraisals of conditions in their specific field. 

(3) Answers to specific questions asked by the chairman. 

A combination of classes (1) and (2) is the most valuable. The 
Committee members should realize that information that may sound 
commonplace to them because of their nearness to the source may be 
of outstanding interest to other branches of the industry. 

Work Preliminary to the Preparation of the Report. The past-chair- 
man of the Committee has found a card file to be the most helpful 
means of coordinating the many hundreds of details which require 
final mention in the report. The contents of this file are assembled 
from three sources, namely: (1) clippings from one or more photo- 
graphic abstract bulletins such as the Monthly Abstract Bulletin of 
the Kodak Research Laboratories, which contains patents as well as 
journal abstracts; (2) abstracts and summaries prepared by com- 
mittee members; (3) miscellaneous data obtained from sources other 
than those mentioned under (1) and (2). One valuable source of 



254 PROGRESS COMMITTEE WORK 

information on trade news is the weekly report of the Motion Picture 
Division of the U. S. Department of Commerce, Washington, D. C. 

From past experience it has been found that illustrations comprise 
a valuable addition to the Progress Report, particularly during its 
presentation. The interest of the audience may be heightened con- 
siderably by the judicious use of lantern slides. A special effort 
should be made to secure illustrations of new equipment developed 
in foreign countries. Short motion picture films of significant develop- 
ments may also be used as a valuable adjunct of the report during its 
presentation, as was shown at the Washington Meeting of the Society 
in May, 1930. 

Preparation of Final Report All reports from the different members 
of the Committee should be in the chairman's hands not later than one 
month before the date of the meeting at which the report will be pre- 
sented. When these are received, every item of value should be card- 
indexed and filed so that, as far as possible, all data to be used in the 
report is on cards. It is not feasible in some cases to transfer the in- 
formation but the reference to it should be prepared so that the data 
may be located with the least possible loss of time. 

When all available data have been filed according to a definite 
classification, the actual writing of the report should be started. All 
references may be made most easily at the time that the material is 
written up, rather than after the writing has been finished. When the 
first rough draft has been typed, the report should be edited for the 
principal items of progress or "highlights" which are to be read at the 
meeting. These highlights should not comprise more than 20 per 
cent of the total report, and sufficient copies (usually about 30) 
should be mimeographed for the use of the Publicity Committee. 

The general introduction to the report giving a broad summary of 
progress should be written last, after a clear impression has been se- 
cured of all the significant developments in the report. Courteous 
acknowledgment should obviously be made to all sources of informa- 
tion and illustrations apart from those actually supplied by Committee 
members. Care should be taken that proper credit is given under 
each illustration published with the report. 

If the work of the Committee is carried out conscientiously and 
thoroughly, this report should become an increasingly valuable 
compendium of technical information on the motion picture industry 
throughout the world. 

G. E. MATTHEWS, Past-Chairman 



ABSTRACTS 

Studio Practice in Noiseless Recording. GEORGE LEWIN. Electronics, Octo- 
ber, 1931, p. 146. The theory of noiseless recording by the light-valve method 
was discussed in a preceding article (Electronics, September, 1931). Some modi- 
fications must be made in adapting the method to studio practice and special in- 
struments must be designed to check the characteristics quickly and accurately. 
The author points out one very practical advantage of noiseless recording 
namely, that the average level may be kept lower, thereby reducing the danger of 
over-shooting. With the introduction of noiseless recording, however, a certain 
amount of background noise that had previously been taken for granted has be- 
come more noticeable. This includes noises originating on the stage or in the 
theater itself due to the ventilating system or projection machines. A. C. H. 

Glow-Lamp Noiseless Recording. E. H. HANSEN. Electronics, November, 
1931, p. 177. A description of the method of producing "noiseless" records by the 
glow-lamp method. A. C. H. 

Ideal Camera Blimp in Daily Use. IRA HOKE. Internal. Phot., 3, November, 
1931, p. 27. A new and extremely useful camera casing is reported from the 
Educational Studios in Hollywood. It is of cast aluminum and sound insulated. 
The new feature is the possibility of pumping the air out with a vacuum pump 
whenever conditions demand the extreme in noiseless equipment. Only 25 
seconds are required in this process and the method interferes in no way with the 
operation of the camera or sound apparatus. A. A. C. 

A Standard Aperture for Sound Films. JOHN ARNOLD. Amer. Cinemat., 12, 
November, 1931, p. 14. Sound on film destroyed the 3X4 proportion of the 
motion picture screen, when it was first introduced. Theaters remedied the con- 
dition by using a reduced aperture of the old proportion, thus forcing the pro- 
ducer to plan his picture to suit, as well as possible, the various sizes that were 
being used in the theaters. This has been accomplished by the expedient of 
masking the camera aperture accordingly, and confining the action to that portion 
of the film. About twenty per cent of the frame area is not used at all, under 
these conditions. 

A new standard, 0.651 X 0.868 inch, for camera aperture and 0.615 X 0.820 inch, 
for projector, is now proposed by the Academy of Motion Picture Arts and 
Sciences. A full report of the proposal is being circulated by Lester Cowan, 
Executive Secretary of the Academy. A. A. C. 

New Photoelectric Cell. Mot. Pict. Proj., 5, November, 1931, p. 37. A de- 
scription is given of the Weston Photronic Cell, which employs a light-sensitive 
disk to transform light directly into electrical energy without the use of auxiliary 
voltage. It delivers about one microampere per foot candle of light intensity 
and the response to light variations is said to be instantaneous. The simplicity 
and ease of operation of the new unit are advantages that are expected to lead to 
its wide use as an indicator in measurements of illumination. A. A. C. 

Rectifying Contact Photoelectric Cells. R. SINGER. Technique Cinemat., 2, 

255 



256 ABSTRACTS [J. S. M. P. E. 

November, 1931, p. 18. It has been known for some time that certain devices, 
notably those using copper oxide in contact with metal for rectifying alternating 
currents, also possessed the property of developing an electrical potential differ- 
ence at their electrodes when radiated with light. The characteristics of two 
commercial cells of this type are discussed briefly. Another cell is mentioned 
which depends on a needle contact with galena crystal. It is stated that these 
cells are rugged and simple in use. They require neither vacuum nor a liquid 
electrolyte. C. E. I. 

How to Determine the Position of the Pick-up Arm. L. LUMIERE. Technique 
Cinemat., 2, November, 1931, p. 4. The author proposes a method of determining 
a position for the pick-up arm which minimizes variation in the angle made with 
the tangent to the record grooves. The geometrical steps are shown. Reference 
is made to an article on this subject which appeared in the preceding issue. 

C. E. I. 

The Panoramic Motion Picture and the Chretien Hypergonar. H. PICARD. 
Technique Cinemat., 2, November, 1931, p. 7. A wide-screen picture can be 
obtained with film of normal width by compressing the image in width by the use 
of an auxiliary cylindrical lens both in making the negative and in projecting the 
positive. This method is open to the objection that the graininess of the negative 
shows up in the magnified image of the positive. It is proposed to overcome this 
fault by using wide negative film and compressing the image by the use of the 
auxiliary lens in the process of projection printing to the fine grain positive. The 
illustrations with the article show pictures of the French Colonial exposition 
buildings made in this manner. Other applications using this scheme are men- 
tioned, such as narrow vertical pictures, and color and stereoscopic processes re- 
quiring two or more pictures in the standard frame. C. E. I. 

New Sound-on-Film Method. Mot. Pict. Herald, 105,October.24, 1931, p. 11. 
This process uses a variable density record on 16-mm. film, having the usual 
double rows of perforations and 40 frames to the running foot of film. The sound 
record is made on a bias which allows greater width of the frequency band, the 
over-all width of the track being 0.025 in. It is claimed to be possible to record 
not only at the old silent speed of 60 feet per minute but also as slowly as 32 feet 
per minute without volume or quality loss. Reduction prints from 35 mm. film 
are planned to form the nucleus of a film library for non- theatrical distribution. 

G. E. M. 

New Photoelectric Cell. Film Daily, 51, November 22, 1931, p. 6. A highly 
light-sensitive disk on the face of this photoelectric cell transforms the light 
energy directly into electrical energy without the use of auxiliary voltage. The 
cell has an instantaneous response to light variations and relays may be operated 
directly from the current generated by the cell. About 1 microampere is delivered 
per foot candle of light intensity. When exposed to direct sunlight, the output is 
about 5 milliamperes. The cell resistance varies from about 1500 ohms for 10 
foot candles to 300 ohms for 240 foot candles. A moulded black bakelite case 
2 x /4 inches in diameter and 1 inch in thickness encloses the cell. G. E. M. 

The Screen: A Problem in Exhibition. BENSCHLANGER. Mot. Pict. Herald, 
105, Sect. 2, October 24, 1931, p. 14. With the exception of the progress made in 
projection engineering, the author claims that the art and science of exhibition 
have advanced very little. The position of the screen, for example, is still being 



Feb., 1932] ABSTRACTS 257 

determined from the stage floor of the drama theater. The average life of a 
theater building should be at least 15 years in order to amortize the initial con- 
struction cost and to show a reasonable investment profit. Bodily comfort of the 
patron is considered of primary importance in theater design. A maximum screen 
size having the ratio of 1 to 1.67 is considered preferable to satisfy various re- 
quirements. G. E. M. 

A Portable Sound Recorder. Kinemat. Weekly, 177, November 19, 1931, p. 
56. A very light and portable sound recording apparatus, capable of being car- 
ried in a small automobile, has been developed by a British manufacturer. The 
recorder may be fitted to almost any modern camera, provided, however, that the 
camera has been silenced for sound work. This comprises changing certain gears 
to fit construction, enclosing the shutter drive in a sound-proof casing, and pro- 
viding more sturdy bearings for the sprockets. 

The recording head and amplifier of this new equipment fit underneath the 
camera in a casing which consists of two compartments; the front chamber carries 
the sound slit and guide rollers while the rear compartment contains a two-valve 
amplifier. The glow lamp projects in front of the forward casing and can be 
slipped out to protect it from damage. The lamp is made of Pyrex glass, and 
special non-spluttering metals are used for the electrodes, thus minimizing the 
risk of the glass turning black. The motor is mounted at the rear of the camera 
case and has incorporated with it a tachometer of improved design. The micro- 
phone used is of the transverse current type. Ear-phones are provided for 
monitoring purposes. C. H. S. 



BOARD OF ABSTRACTORS 

BROWNELL, C. E. MACFARLANE, J. W. 

CARRIGAN, J. B. MACNAIR, W. A. 

COOK, A. A. MATTHEWS, G. E. 

CRABTREE, J. I. McNicoL, D. 

FOWELL, F. MEULENDYKE, C. E. 

HAAK, A. H. MUEHLER, L. E. 

HARDY, A. C. PARKER, H. 

HERRIOT, W. SANDVICK, O. 

IRBY, F. S. SCHWINGEL, C. H. 

IVES, C. E. SEYMOUR, M. W. 

LOVELAND, R. P. WEYERTS, W. 



ABSTRACTS OF RECENT U. S. PATENTS 

The views of the readers of the JOURNAL relative to the usefulness to them of the 
patent abstracts regularly published in the JOURNAL will be appreciated. Favorable 
views are of particular interest. In the absence of a substantial body of opinion to the 
effect that these patent abstracts are desired by the membership, their early discon- 
tinuance may be considered. 

1,825,663. Film Reel and Spindle. A. G. HAYDEN. Oct. 6, 1931. A reel 
and spindle interlock on slight relative rotative adjustment, thereby to give them 
a driving connection and prevent the reel from accidental escape from the spindle. 
The film reel comprises a pair of plates and a hub between said plates adapted to 
have a film wound thereon, one of said plates having a hole therein and the other 
of said plates having an opening with tongues therein projecting toward, but not 
to, the center of the plate; and a spindle, in said hole and opening, having a groove 
for receiving said tongues to prevent movement of the reel axially of the spindle. 

1,825,781. Television Scanning Device. L. H. DAWSON. Oct. 6, 1931. 
Scanning disk for television systems in which a rotatable disk is provided with a 
plurality of conically shaped light conducting and concentrating members extend- 
ing through the disk perpendicularly to the plane thereof. The light concentrat- 
ing members are constructed from quartz having a high refractive index for in- 
creasing the luminous intensity of the image by concentration of the available 
light rays. 

1 ,825,953. Device for Permitting the Continuous Feeding of the Film in Project- 
ing Apparatus. P. G. H. HALLONGREN. Oct. 6, 1931. Projecting apparatus in 
which the reflecting members are divided into at least two groups, which suc- 
cessively reflect the picture rays and are positively caused to turn synchronously, 
during which operation the active surfaces or the surfaces struck by the picture 
rays turn in the same direction, and the said rays pass the reflecting surfaces at 
the same side of the axis or axes of rotation through which the said reflecting 
surface or surfaces extend or with which the surfaces or surface are substantially 
parallel, the said axis or axes having an oblique position with relation to the 
plane, on which the incoming rays travel (the plane of the wandering picture). 

If two axes of rotation are provided the reflecting surfaces may be located 
either round the axes or tangentially to cylindrical surfaces enclosing the axes of 
rotation and concentric with the same. In practice the two groups of reflecting 
members preferably are located around an axis common to the same and the re- 
flecting surfaces of the one group located radially, while the reflecting surfaces of 
the second group are located tangentially to a cylindrical surface enclosing the 
said axis and concentric with the same. 

1,825,955. Synchronized Cylinder Record for Talking Picture. E. S. HAY- 
FORD. Oct. 6, 1931. Apparatus for synchronizing a sound record with a picture 
record comprising a cylinder of conducting material mounted for simultaneous 
movement with the picture record, a sleeve of non-conducting material carried 
258 



PATENT ABSTRACTS 259 

upon said cylinder and having an opening, a stylus mounted for movement along 
said cylinder and normally engaging said sleeve and adapted to enter the opening 
therein and an electrically operated actuating device connected in circuit with 
said stylus and said cylinder for operating said sound record. The sound pro- 
ducing means may be rendered operative or inoperative at any predetermined 
position with respect to the film being projected, thereby permitting the use of a 
record having a limited tone groove length in connection with a greater length of 
film. 

1,826,305. Scanning System for Television. H. P. DONLE. Oct. 6, 1931. A 
scanning system having a speed regulating drive interposed between the scanning 
disk and the driving motor. The shaft is formed in two parts, and the speed of 
rotation of one part is manually controlled by friction means and regularized by 
a ball governor. The other part of the shaft carries the scanning disk and the 
angular relation between the two parts of the shaft is adjustable by manually 
controlled means independent of the speed controlling means and independent of 
the speed of the motor. 

1,826,332. Drive Mechanism for Scanning Disk. C. O. VERMILLION. As- 
signed to Wired Radio, Inc. Oct. 6, 1931. A drive mechanism for a scanning 
disk having means for framing the scanning holes of the scanning disk with respect 
to the object to be televised or the picture to be reproduced. The scanning disk 
driving mechanism is so arranged that constant speed may be obtained at both 
the transmitter and receiver even during periods of adjustment for framing the 
apertures in the scanning disk with respect to the picture or object. 

1,826,522. System for Avoiding Interruptions of Television Program. F. H. 
OWENS. Assigned to Owens Development Corp. Oct. 6, 1931. A plurality of 
photoelectric cells are arranged in light paths formed through the film. The cells 
operate simultaneously for controlling the input circuit of an amplifying system. 
The light which is directed through the film is split into diverging paths toward a 
plurality of photoelectric cells so that any one of the cells will continue to operate 
for controlling the reproduction of sound in the event of failure of the others so 
that there will be no interruption to the sound program. 

1,826,680. Motion Picture Projector Cabinet. A. STUBER. Assigned to 
Eastman Kodak Co. Oct. 6, 1931. A projector is housed with a sound repro- 
ducing instrument in the same cabinet, the projector being mounted on a rota table 
support for projecting a picture in any desired direction to the most suitable 
location on a portable screen. A phonograph or radio apparatus may be housed 
in the cabinet, but is so isolated from the projector that the noises of the projector 
are muffled and prevented from interfering with the equipment within the cabinet. 
The light rays from the projector within the cabinet are directed vertically through 
the cabinet and then projected horizontally in any desired angular direction. The 
direction of the beam may be selected by shifting the projector to the desired 
angular position within the cabinet structure by means of a crank which engages 
the rotatable mount for the projector. 

1,826,695. Light-Protected Motion Picture Film. P. FAVOUR. Assigned to 
Eastman Kodak Co. Oct. 6, 1931. A light protecting covering is interwound 
with the film strip and is normally unperforated, but capable of being perforated 
as a film moving mechanism advances the film through contact with the film 
perforations. Pasters are provided for attaching the supplementary light- 



260 PATENT ABSTRACTS [J. S. M. P. E. 

protective covering to the perforated film band, the pasters attaching one end 
only of each supplementary light-protective covering to the film band. 

1,826,754. Method of Making Photophonographic Records. F. EHRENHAFT. 
Oct. 13, 1931. A recording lamp is employed having a luminescent gas discharge 
controlled by sound waves, which transform said luminescent gas discharge into 
a transitional form of discharge intermediate between a glow and arc discharge. 

1,826,786. Sound Controlled Still Picture Protector. P. S.HOPKINS. Assigned 
by mesne assignments to Agfa Ansco Corp. Oct. 13, 1931. Projecting appara- 
tus for still pictures accompanied by a sound program. The still pictures are 
shifted automatically to coordinate the picture with the sound program so that a 
picture is projected appropriate to the accompanying sound. The apparatus is 
capable of use as a projector accompanied by an illustrated lecture without the 
attendance of the lecturer. 

1,826,812. Electrooptical Transmission Employing Mirrors instead of Light 
Valve. H. NYQUIST. Assigned to American Tel. and Tel. Co. Oct. 13, 1931. 
A system for transmitting electrical impulses into light impulses of varying in- 
tensities, comprising two plane mirrors having their planes intersecting at right 
angles and controlled by incoming picture current at the receiving station, which 
mirrors take the place of the usual light valve. The term "90-degree mirror" is 
used to designate such an arrangement of plane mirrors. This "90-degree 
mirror" rotates about an axis at the line of intersection. The surfaces consist of 
alternately reflecting and non-reflecting strips which gradually increase in width 
from the line of intersection. The rotation of the 90-degree mirror is controlled 
jointly by picture currents received from a transmission line, which currents pass 
through a movable coil attached to the 90-degree mirror, and by current from a 
local source which passes through a stationary coil, the position of the 90-degree 
mirror varying in accordance with the amount of current received from the line. 
A constant light source is arranged to project a beam of substantially parallel 
rays of light toward the surfaces, the axis of the beam being directed toward the 
axis, or intersection line, of the surfaces and at an angle thereto. The reflected 
beam from these surfaces is directed to a focal point on a light-sensitive surface, 
such as a photographic film. The amount of light reflected by the surfaces and, 
therefore, the intensity of the light at the focal point will vary directly with the 
angular change in position of the surfaces as controlled by the picture currents 
received from the sending station. The reflecting strips on the surfaces may be 
so designed as to give a non-linear relation between the light intensity and the 
received current strength. 

1,826,836. Television Scanning Device. M. STACHO. Oct. 13, 1931. A 
television scanning system consisting of a pair of rotatably mounted disks having 
co-acting intersecting slots therein for the passage of light rays. One of the disks 
has an armature member mounted thereon and associated with an electromag- 
netic control for retarding the disk at the completion of each revolution in a 
manner to cause the same to rotate periodically at a reduced speed as compared 
with the other disk. 

1,826,858. Photographic printing apparatus. V. K. ZWORYKIN. Assigned to 
Westinghouse Electric and Manufacturing Co. Oct. 13, 1931. A concentric 
arrangement of drums for aligning a positive film with a negative film for the 
printing of positives from the negative. The light source is directed through the 



Feb., 1932] PATENT ABSTRACTS 261 

drums and through the negative film adjacent to the outside drum to the positive 
film adjacent to the inner drum. The light source, when a reduction in film size 
is to be made, is positioned exteriorly of the large wheels over which the negative 
film is fed, and the light therefrom, passing through the negative film, falls upon 
the surface of the unexposed film carried over the smaller wheels. If the device 
is to be used for enlarging, the negative film is fed across the small wheels and the 
positive film across the large wheels, the light source being so re-positioned that 
the negative film passes between it and the positive film. 

1,826,970. Television and Telephoto Device. J. L. WALKER. Oct. 13, 1931. 
Picture reproducing system in which two separate scanning systems are directed 
upon opposite sides of a reproducing screen. A photographic plate or viewing 
screen uses light from two separate light sources and projects light from one light 
source upon one side and from the other light source upon the other side of said 
photographic plate or viewing screen and the illumination from the two separate 
light sources combined at one point. The recording lamps of the two scanning 
systems are connected in parallel in the output circuit of the receiving apparatus, 
and each so positioned on opposite sides of the screen as normally to give equal 
illumination upon the screen. 

1,827,010. Film Flame Stop. L. D. KOHLMEYER. Oct. 13, 1931. The film 
is protected by a fire-proof frame structure forming compartments surrounding 
the film reels. The entrances to each of the compartments are provided with 
passageways formed between a pair of rollers carried on fixed axes in the passage- 
way. A second pair of rollers is mounted adjacent to each passageway for guid- 
ing the film through the passageway and at the same time forming a fire stop in 
the event of ignition of the film. 

1,827,018. Photoelectric Cell. A. JOFFE. Assigned to Industrial Research 
Co. Oct. 13, 1931. A photoelectric cell comprising a sheet-like insulating 
layer having a thickness not greater than 0.01 mm. having a photoelectrically 
active substance distributed through the insulating layer and a pair of electrodes 
supporting the layer, at least one of the electrodes being transparent to light. 
The invention is based on the discovery that when an ion is initiated or excited 
within certain substances of requisite thickness, notably dielectrics or other ma- 
terials of low specific conductivity, and further, when the substance is subjected 
to considerable electrical stress, the medium through which the ion travels at 
high velocity gives rise to an augmentation of the number of charged particles. 
The accumulative action effects a general movement of ions toward one of the 
electrodes and results in a greatly magnified space current with abrupt reduction 
of impedance to produce amplification of the impulse originally exciting the 
single ion. The original impulse may be energy derived from any physical phe- 
nomenon such as light, heat, electron bombardment, or other electrical effects. 

1,827,206. Film Support for Photographic Apparatus. F. H. OWENS. As- 
signed to Owens Development Corp. Oct. 13, 1931. A support for traveling 
films, comprising a pair of axially aligned movable members, one of said members 
being adapted to engage a film and cause the same to travel over the other mem- 
ber. A stationary member is disposed between said movable members and 
spaced therefrom to permit the passage of light to said film between said stationary 
and movable members and on each side of said stationary member. 

1,827,282. System of Composite Photography for Motion Pictures. O. 



262 PATENT ABSTRACTS [j. s. M. P. E. 

CHOUINARD. Assigned to Motion Picture Improvements, Inc. Oct. 13, 1931. 
A machine to produce moving pictures of animated objects and scenic or other 
effects wherein the scenic or other effects are recorded in positive, direct, and 
accurate relation to the moving objects, without the heavy cost of "locating." 
The method comprises making duplicate exposures on two films of moving objects 
having actinic properties substantially different from those of the background 
therefor, developing one of said films, projecting images from the respective 
frames of said developed film successively toward an actinic background, suc- 
cessively altering the actinic effect of said background complementary to and in 
registration with the respective projected images, and doubly exposing said un- 
developed film by subjecting its respective frames to said background as suc- 
cessively altered in actinic effect and without substantial effect thereon of the 
respective projected images. 

1,827,588. Film Drive. E. W. KELLOGG. Assigned to General Electric Co. 
Oct. 13, 1931. An improved film driving apparatus in which the film is driven 
jointly by a sprocket and a roller or drum and in which the speed of one of said 
members is varied in accordance with the amount of film moved by the respective 
members as determined by the number of film sprocket holes. A free running 
sprocket hole counter is provided engaging that portion of the film moved by the 
drum and a variable speed driving mechanism for the drum controlled by the rela- 
tive movement of a drive sprocket and the sprocket hole counter. There are 
means responsive to a difference in speed of those portions of the film moved by 
the respective sprocket and drum members, as determined by the sprocket tooth 
openings and independently of the length of film between said members, for vary- 
ing the speed of one of said members. 

1,827,598. Motion Picture Cabinet. A. G. MERRIMAN. Oct. 13, 1931. The 
projecting apparatus is mounted within a cabinet structure having a portion at 
one side thereof which may be moved away from the cabinet structure for sup- 
porting a projecting screen upon which the picture from the projecting apparatus 
within the cabinet structure may be displayed. When the apparatus is not in 
use the screen is foldable into a position within the cabinet structure, making a 
compact article of furniture for the home or a compact advertising apparatus. 

1,827,735. Volume Control in Sound Record Reproduction. J. R. BALSLEY. 
Assigned to Fox Film Corp. Oct. 13, 1931. The film bearing the sound record 
also carries a volume control record driven in synchronism with the sound 
record, and adapted to control the volume level of the sound reproduced from the 
sound record. This volume control record may be simply a varying density 
photographic record, which may be prepared by reference to the volume level of 
the sound record as recorded, as may be determined by ordinary reproduction 
thereof. The volume control record, which may be printed on the same film that 
carries the pictures and sound record, for instance, outside the sprocket perfora- 
tions thereof, or on a separate film if more convenient, is operated in conjunction 
with a light beam and photoelectric cell to produce a varying electrical current 
which is utilized to control the level of reproduction, and to do this irrespective 
of the level at which the sound record was recorded. The photoelectric cell 
which is acted upon by the volume control record is connected across the grid and 
plate of a vacuum tube, whereby a varying plate current corresponding thereto 
appears in the plate circuit of the tube with means for modifying the volume 



Feb., 1932] PATENT ABSTRACTS 263 

level of the reproduced sound in accordance with the variations in said plate 
current. 

1,827,924. Picture Copying Process. F. D. WILLIAMS. Oct. 20, 1931. A 
method of copying pictures which comprises projecting primary component sil- 
houette pictures of ultimate composite pictures upon an opaque picture perceptive 
screen and light-impressing a sensitized medium with a supplementary compo- 
nent, by aid of the light from said screen with the silhouette projected thereon 
so as to produce a latent stencil area. The stencil area is then light-impressed 
with a regular picture corresponding to the silhouette. 

1,827,947. Synchronizing Mechanism for Disk Reproduction. W. R. MOORE, 
JR. Assigned to Deca Disk Phonograph Co. .Oct. 20, 1931. Mechanical link- 
age for connecting phonograph and a picture projecting machine for taking up 
all lost motion between the mechanism for playing the record and that for pro- 
jecting the pictures so that the music and the pictures shall perfectly synchronize. 
A worm gear connection is provided with an adjusting device which permits the 
taking up of lost motion. 

1,828,032. Projection Machine with Optical Intermittent. R. DECAUX. 
Assigned to SocietS des 6tablissements Gaumont. Oct. 20, 1931. Projector 
wherein the film moves in a continuous manner along an arcuate guide, past a 
window lighted by a luminous source which is combined with a condenser. The 
film occupies the focal plane of an optical system which sends a beam of parallel 
rays' on a mirror which is caused to oscillate about an axis located in its plane. 
From that mirror, the luminous rays are directed on a stationary mirror disposed 
at 45 degrees, caused to pass through an objective, from which they are projected 
on the screen. The oscillating movement of the mirror, which is controlled by a 
cam, is synchronized with the forward movement of the band in such a way that, 
between successive extinctions produced by a rotary blade acting as a shutter, the 
image of a determined point on the film is maintained stationary on the screen. 
The chief object of the invention is to provide a mechanical arrangement of the 
parts owing to which the oscillating mirror, the support of said mirror, and the 
control cam for controlling it are caused to cooperate under the best conditions, 
account being taken of the inertia of the different pieces and of the play which is 
liable to take place as a consequence of wear and tear. The mirror is fixed on a 
platform pivoted to a rocking lever of adjustable position and carrying an arm 
which receives the oscillations of the cam. The mirror bears at three points on 
the platform and is maintained in place by springs, in such a way as to eliminate 
all deformation of the reflecting surface. 

1,828,199. Toy Talking Movie Device. F. H. OWENS. Oct. 20, 1931. An 
inexpensive form of toy talking picture apparatus wherein an intermittent picture 
strip may be moved past a viewing window in timed relation to the movement of 
a rotatable talking machine record support. The record carries the sound ap- 
propriate to the picture and is maintained at proper operating speed by a governor 
device. 

1,828,236. Method of Producing Visual Effects. A. C. WATSON. Oct. 20, 
1931. A neon lamp illuminating device in which substantially instantaneous in- 
termittent illuminations are formed in different positions along a periodic path in 
rapid succession through repetitive cycles satisfying the critical frequency for 
continuous visual sensation. Visual effects of appreciable duration are produced 



264 PATENT ABSTRACTS [J. S. M. P. E. 

and modified by interposing a mask between the illuminations and the observer. 
An instance of usefulness of this method consists in the fact that by combining 
the red color of neon with the yellow color obtained from it as in the "Bezold- 
Brucke" phenomenon and also with other types of light such as the neon mercury 
tube and by placing before the rotating light a rotating mask which may itself be 
colored, so as to reflect daylight, it is possible to secure vari-colored visual patterns. 
If the mask referred to be rotated at a slightly different speed from that of the 
light, then the colored patterns undergo a series of changes of form, as well as of 
color and the total effect may be upon such a large scale as to produce exceedingly 
attractive and beautiful patterns of various colors. 

1,828,364. Film Contact System Employing Air Pressure. F. E. GARBUTT. 
Assigned to Paramount Publix Corp. Oct. 20, 1931. The positive and negative 
films are pressed into firm contact by an air pressure system in connection with 
the printer and a current of air directed against the films in such a manner that 
the films are held in perfect contact against the registering means upon which 
they are supported. 

1,828,399. Photoelectric Cell Light Ray Condenser. C. W. EBELING. As- 
signed to General Talking Pictures Corp. Oct. 20, 1931. A photoelectric cell 
light ray condenser is provided for condensing the rays of light after the same 
have passed through the sound track of the film and before the same impinges 
upon the photoelectric cell, thus insuring higher efficiency in the action from the 
cell due to the concentration of the beam of light thereon. A condensing lens is 
carried in the light slit block in the path of the light rays before they reach the 
photoelectric cell. 

1,828,444. Method of Dubbing and Printing. W. ROM. Oct. 20, 1931. A 
printer for applying a sound record to a previously prepared picture film, which 
consists in utilizing two positive films of the same picture and projecting one 
positive film on a screen for guidance in applying sound to a negative film made 
from the other positive film of the same picture, driving said other positive of said 
film in synchronism with the projected film, masking a portion of said other 
positive thereby to provide an area for the sound record, driving a negative film 
in synchronism and printing relation with said other positive and with the sound 
area of said other positive masked as to said negative, and simultaneously record- 
ing sound on the sound area of said new negative, the sound record being applied 
to the sound area of said negative in accordance with the projected positive of the 
same picture. 

1,828,569. Film Stopping Apparatus. E. W. KELLOGG. Assigned to General 
Electric Co. Oct. 20, 1931. The projector is arranged to stop the film driving 
machine before the record film is completely unwound and disengaged from the 
reel on which it has been wound. This is the situation, for example, when in 
normal operation the film is rewound on the original reel without removal from 
the machine, the purpose of rewinding being to leave the film ready for immediate 
use, namely, with the beginning part of the record on the outside. 

1,828,571. Picture Transmission System. I. LANGMUIR. Assigned to Gen- 
eral Electric Co. Oct. 20, 1931. A light source of the flaming arc type is used 
at the picture receiver. The current supplied to the arc lamp is modulated in 
accordance with the received signal. The picture at the receiver is projected on 
a screen. Spots of light from the arc lamp are projected on the screen but light 



Feb., 1932] PATENT ABSTRACTS 265 

from the electrodes excluded. This is accomplished by a scanning apparatus 
comprising a disk having a series of lenses arranged in a spiral therein and ar- 
ranged successively to pass between the lamp and the screen when the disk is 
rotated with a motor for rotating the disk in synchronism with a sending appara- 
tus. An objective lens is provided and a second disk rotatable with the first- 
mentioned disk arranged with a series of holes therein corresponding with said 
lenses for excluding from the objective all light emanating from the electrodes of 
the lamp. 

(Abstracts compiled by John B. Brady, Patent Attorney, Washington, D. C.) 



BOOK REVIEWS 



Handbook of the Film Industry, Vol. II, European Films (Handbuch der 
Filmwirtschaft, Band II, Film-Europa). Wirtschaft und Politik, Berlin, 1931, 
272 pp. 

Three volumes of this handbook of film statistics have thus far appeared. The 
first volume covered the period 1923 to 1925, giving a cross-indexed register of 
information of film productions, authors, directors, cameramen, architects, and 
producers. The history and development of the German motion picture pro- 
ducing and exhibiting industry were also traced from 1895 to 1923, together with 
an outline of the general film situation in Europe. 

The second volume gives correspondingly indexed statistics for films produced 
and passed by the censors during 1926 to 1929 with indexes of authors, etc. 
Statistics also give information as to the size and distribution of theaters in the 
various countries of Europe, regulations pertaining to the importation of motion 
picture productions into these countries, the general film market in Europe, 
division of sales, etc. The book will be of greatest use to executives, film sales 
and distributing organizations doing business with Europe. 

A third volume of Handbuch der Filmwirtschaft, dealing with the rise of the 
sound film industry and covering the period 1929 and 1930, is scheduled to appear 
during 1931. 

L. E. MUEHLER 

Sound Film Reproduction. G. F. JONES. Blackie & Son, Ltd., London & 
Glasgow. 1931. 

A brief text in simple, non- technical style explaining, primarily for the small 
theater manager and projectionist, the principles and details "of construction of 
reproduction equipment for both disk and sound-on-film. The principal outfits 
available on the British market are described. Sections are devoted to the vari- 
ous parts of the equipment as turntables, pickups, sound heads, light-sensitive 
cells, amplifiers, etc. A section on home-designed installations mentions the 
chief problems to be met but points out that very little saving can be effected by 
such assemblies. 

H. PARKER 



266 



SOCIETY OF MOTION PICTURE 
ENGINEERS 

OFFICERS 
1931-1932 

President 
A. N. GOLDSMITH, Radio Corporation of America, New York, N. Y. 

Past-President 
J. I. CRABTREE, Eastman Kodak Company, Rochester, N. Y. 

Vice-Presidents 

W. C. HUBBARD, General Electric Vapor Lamp Co., Hoboken, N. J. 
E. I. SPONABLE. Fox Film Corp., New York, N. Y. 

Secretary 
J. H. KURLANDER. Westinghouse Lamp Co., Bloomfield, N. J. 

Treasurer 
H. T. COWLING, Eastman Teaching Films, Inc., Rochester, N. Y. 

Board of Governors 

F. C. BADGLEY, Canadian Government Motion Picture Bureau, Ottawa, Canada 
H. T. COWLING, Eastman Teaching Films, Inc., 343 State St., Rochester, N. Y. 
J. I. CRABTREE, Research Laboratories, Eastman Kodak Co., Rochester, N. Y. 
P. H. EVANS, Warner Bros. Pictures, Inc., 1277 E. 14th St., Brooklyn, N. Y. 
O. M. GLUNT, Bell Telephone Laboratories, New York, N. Y. 
A. N. GOLDSMITH, Radio Corporation of America, 570 Lexington Ave., New 

York, N. Y. 

W. C. HUBBARD, General Electric Vapor Lamp Co., 'Hoboken, N. J. 
R. F. MITCHELL, Bell & Howell Co., 1801 Larchmont Ave., Chicago, 111. 
J. H. KURLANDER, Westinghouse Lamp Co. Bloomfield, N. J. 
W. C. KUNZMANN, National Carbon Co., Cleveland, Ohio 

D. MACKENZIE, Electrical Research Products, Inc., 7046 Hollywood Blvd., 

Los Angeles, Calif. 
L. C. PORTER, General Electric Co., Nela Park, Cleveland, Ohio 

E. I. SPONABLE, 277 Park Ave., New York. N. Y. 

267 



268 



COMMITTEES 



[J. S M. P. E 



COMMITTEES 
1931-1932 

(The completed list of committees will be published in a later issue] 



W. C. HUBBARD 



Convention 
W. C. KUNZMANN, Chairman 



M. W. PALMER 



J. I. CRABTREE 
E. J. DENISON 
T. FAULKNER 



Development and Care of Film 
R. F. NICHOLSON, Chairman 
R. C. HUBBARD 

K. MAClLVAIN 

D. MACKENZIE 



J. S. MACLEOD 
H. RUBIN 
J. H. SPRAY 



H. T. COWLING 
W. B. COOK 



Finance 

L. A. JONES, Chairman 
J. I. CRABTREE 
W. C. HUBBARD 



J. H. KURLANDER 

L. C. PORTER 



W. CLARK 



Historical 
C. L. GREGORY, Chairman 



N. D. GOLDEN 



O. M. GLUNT 



Journal and Progress Medal Awards 
C. E. K. MEES, Chairman 

E. A. WILLIFORD 



Membership and Subscription 

H. T. COWLING, Chairman 
W. H. CARSON, Vice-Chairman 

D. M. BALTIMORE C. D. ELMS J. KLENKE 
J. R. CAMERON R. EVANS E. E. LAMB 

E. J. COUR E. R. GEIB T. NAGASE 

B. W. DEPUE 'J. G. T. GILMOUR E. C. SCHMITZ 



B. W. DEPUE 
O. B. DEPUE 

C, L. GREGORY 



Museum 

W. E. THEISEN, Chairman 
C. F. JENKINS 
F, H, RICHARDSON 



T. RAMSAYE 
A. REEVES 
A. F. VICTOR 



Feb., 1932] 



COMMITTEES 



269 



Non-Theatrical Equipment 
R. E. FARNHAM, Chairman 

A. A. COOK N. B. GREEN 

W. B. COOK H. GRIFFIN 

R. F. MITCHELL 



A. SHAPIRO 
A. F. VICTOR 



J. A. BALL 
C. DREHER 
P. H. EVANS 
A. C. HARDY 
N. M. LA PORTE 



Papers 
O. M. GLUNT, Chairman 

G. E. MATTHEWS 
P. A. McGuiRE 
G. A. MITCHELL 
D. McNicoL 



P. MOLE 
K. F. MORGAN 
C. N. REIFSTECK 
P. H. REISMAN 
T. E. SHEA 



H. T. COWLING 
J. I. CRABTREE 



Preservation of Film 
W. H. CARSON, Chairman 
A. S. DICKINSON 
R. EVANS 
C. L. GREGORY 



T. RAMSAYE 
V. B. SEASE 



G. A. CHAMBERS 
C. DREHER 
W. C. HARCUS 



Progress 

J. G. FRAYNE, Chairman 
G. E. MATTHEWS 
M. W. PALMER 



G. F. RACKETT 
H. SINTZENICH 
S. K. WOLF 



J. O. BAKER 
T. BARROWS 
W. H. BELTZ 
G. C. EDWARDS 
S. GLAUBER 



Projection Practice 
H. RUBIN, Chairman 
J. H. GOLDBERG 
C. GREENE 
H. GRIFFIN 
J. HOPKINS 

R. H. MCCULLOUGH 

P. A. McGuiRE 



R. MlEHLING 

F. H. RICHARDSON 
M. RUBEN 
P. T. SHERIDAN 
L. M. TOWNSEND 



J. L. CASS 
H. GRIFFIN 

J. H. KURLANDER 



Projection Screens 

S. K. WOLF, Chairman 
W. F. LITTLE 
A. L. RAVEN 



H. RUBIN 

L. T. TROLAND 

C. TUTTLE 



L. DEL RlCCIO 



Projection Theory 
A. C. HARDY, Chairman 



W. F. LITTLE 



270 



COMMITTEES 



F. C. BADGLEY 
B. W. DEPUE 



Publicity 

W. WHITMORE, Chairman 
D. E. HYNDMAN 
F. S. IRBY 



G. E. MATTHEWS 
D. McNicoL 



M. C. BATSEL 
P. H. EVANS 
N. M. LA PORTE 



Sound 

H. B. SANTEE, Chairman 
E. W. KELLOGG 
C. L. LOOTENS 
W. C. MILLER 



H. C. SILENT 
R. V. TERRY 
S. K. WOLF 



L. E. CLARK 
L. DE FOREST 

J. A. DUBRAY 

P. H. EVANS 
R. E. FARNHAM 
H. GRIFFIN 
A. C. HARDY 



L. J. BUTTOLPH 
R. E. FARNHAM 



Standards and Nomenclature 
M. C. BATSEL, Chairman 
R. C. HUBBARD 
L. A. JONES 
N. M. LA PORTE 
D. MACKENZIE 
G. A. MITCHELL 
G. F. RACKETT 

Studio Lighting 
M. W. PALMER, Chairman 
C. W. HANDLEY 
K. C. D. HICKMAN 



W. B. RAYTON 
C. N. REIFSTECK 
V. B. SEASE 
T. E. SHEA 
J. L. SPENCE 
E. I. SPONABLE 
L. T. TROLAND 



J. H. KURLANDER 

E. C. RICHARDSON 



R. S. BURNAP 
W. H. CARSON 



Ways and Means 
D. McNicoL, Chairman 
H. GRIFFIN 
F. S. IRBY 



J. H. KURLANDIiR 
J. A. NORLING 



Chicago Section 

R. F. MITCHELL, Chairman R. P. BURNS, Manager 

B. W. DEPUE, Sec.-Treas. O. B. DEPUE, Manager 

New York Section 

P. H. EVANS, Chairman M. C. BATSEL, Manager 

D. E. HYNDMAN, Sec.-Treas. J. L. SPENCE, Manager 



Pacific Coast Section 

D. MACKENZIE, Chairman C. DREHER, Manager 

W. C. HARCUS, Sec.-Treas. H. C. SILENT, Manager 



CONTRIBUTORS TO THIS ISSUE 

Frederick, H. A.: B.S., E.E., Princeton University; engineering department, 
Western Electric Company, 1912-25; transmission instruments director, Bell 
Telephone Laboratories, 1925 to date. 

Schlanger, B.: See August, 1931, issue of JOURNAL. 

Sheppard, S. E.: Born 1882 at Hither Green, Kent, England. D.Sc., Uni- 
versity of London, 1906; colloid chemist, Eastman Kodak Company, 1913-26; 
chief of department of physical and inorganic chemistry, 1920; acting director 
of research, 1922-23; assistant director of research, 1924 to date. 

Tuttle, C.: Born March 7, 1898, at Evansville, Wis. B.A., University of 
Wisconsin, 1922; graduate assistant at University of Wisconsin, 1922-23; 
instructor in physics, University of Georgia, 1923-24; physicist, Kodak Research 
Laboratories, Eastman Kodak Company, 1924 to date. 

Tuttle, W. N.: A.B., Harvard University, 1924; S.M. in electric communica- 
tion engineering, Harvard University, 1926; Ph.D., Harvard University, 1929; 
instructor in physics, Harvard University, 1929-30; engineer, General Radio 
Company, 1930 to date. 



271 



SOCIETY ANNOUNCEMENTS 

SPRING, 1932, MEETING 

By action of the Board of Governors at a meeting held on Decem- 
ber 10th at New York, N. Y., and subsequent verification of this 
action by the post-card ballot mailed to the membership for de- 
termining the location of the Spring, 1932, Meeting, the location of 
the latter was determined as Washington, D. C. 

The meeting is to be held from May 9th to 12th, inclusive, with 
headquarters at the Wardman Park Hotel, in Washington. The 
technical meetings will be held in the Little Theater of the Hotel and 
the semi-annual banquet in the Gold Room. The Convention 
Arrangements Committee, under the chairmanship of Mr. W. C. 
Kunzmann, is working on an attractive and interesting program for 
the Convention, and the Papers Committee, headed by Mr. O. M. 
Glunt, is bending every effort toward making the technical sessions 
of outstanding interest. 

Mr. N. D. Golden, of Washington, has been appointed Chairman 
of the Local Arrangements Committee, and in this capacity is 
assisted by Messrs. C. Francis Jenkins, Raymond Evans, C. N. 
Nichols, N. Glasser, C. J. North, and N. C. Haefele. As the Con- 
vention is to be held at the time of the Washington Bi-Centennial, 
there will be much in Washington to attract the members of the 
Society to the Convention, in addition to the technical activities of 
the Society. 

An exhibit will be held of newly developed motion picture appara- 
tus, similar to the exhibits held at the Hollywood and Swampscott 
Conventions. This exhibit is to be under the direction of Mr. H. 
Griffin. As in the past, it will not be of the nature of a 
trade exhibit nor will there be booths, but adequate space will be 
allotted each exhibitor free of charge. The exhibition rules specify 
that equipment be new or have been improved within the past 
twelve months. No pamphlets or advertising literature will be 
permitted. Each exhibitor will be allowed to display a small card 
giving the name of the manufacturing concern, and each piece of 
equipment will be labelled with a plain label free of the name of the 
272 



SOCIETY ANNOUNCEMENTS 273 

manufacturer. It is required that a technical expert be present 
during the exhibition to explain the technical features of the ap- 
paratus. 

Requests for space should be made to Mr. Sylvan Harris, editor- 
manager of the Society, room 701, 33 W. 42nd Street, New York, 
N. Y., stating the number and nature of the items to be exhibited. 

STANDARDS COMMITTEE 

At a meeting of the Standards and Nomenclature Committee, 
held at the General Office of the Society on January 9th, Mr. A. C. 
Hardy was appointed chairman of the sub-committee on the glos- 
sary. Questions on the standardization of camera motors, aper- 
tures, camera mountings, and adapters were discussed, and general 
ideas concerning the disposition of these matters were outlined. 
The proposal made by the Projection Practice Committee, calling 
for the dimensions 0.590 X 0.825 inch for the projector aperture, 
was approved by the Committee. 

The question of standardization of screen brightness was given 
considerable study, and it was finally agreed that the Projection 
Practice Committee, in collaboration with the Projection Screens 
Committee, should study the problem and recommend to the Stand- 
ards Committee, the values of brightness which will indicate the 
limits between which a picture may be considered reasonably satis- 
factory under existing practical conditions. These values would 
not be susceptible of standardization, but would merely represent 
recommended good practice. 

Mr. L. A. Jones was appointed chairman of a new sub-committee 
on sensitometry, and Mr. J. L. Spence was appointed chairman of 
the sub-committee to deal, with matters relating to the standardiza- 
tion of 16-mm. sound-on-film equipment. Considerable thought was 
given to the dimensions of the 16-mm. sound film and the location 
of the sound track, etc., and plans were made for enlisting the assis- 
tance of the manufacturers in making the study, which requires a 
practical knowledge of possible tolerances and practical circum- 
stances of manufacture. 

In connection with the questions raised by the Cine-Standards 
Committee of the International Congress of Photography, recently 
held in Dresden, Germany, it was decided that the matter of specifi- 
cations of safety film is to be reopened at a later meeting of the 
Committee. 



274 SOCIETY ANNOUNCEMENTS [J. S. M. p. E. 

WAYS AND MEANS COMMITTEE 

At a meeting of the Ways and Means Committee, held at New 
York on January 9th, under the chairmanship of Mr. D. McNicol, 
many of the factors involved in readjusting the rates of subscription 
for the JOURNAL, the entrance fees of new members, and dues were 
discussed, and recommendations concerning the reduction of the 
entrance fees and subscription rates were agreed upon for presenta- 
tion to the Board of Governors. 



JOURNAL BINDERS 

The binder shown in the accompanying illustration serves as a 
temporary transfer binder or as a permanent cover for a complete 
year's supply of JOURNALS. It is made of black crush fabrikoid, 
with lettering in gold. The binder is so constructed that each in- 
dividual copy of the JOURNAL will lie flat as its pages are turned. 
The separate copies are held rigidly in place but may be removed or 
replaced at will in a few seconds. 




These binders may be obtained by sending your order to the 
General Office of the Society, 33 West 42nd Street, New York, N. Y., 
accompanied by a remittance of two dollars. Your name and the 
volume number of the JOURNAL may be lettered in gold on embossed 
bars provided for the purpose at a charge of fifty cents each. 



Feb., 1932] SOCIETY ANNOUNCEMENTS 275 

MEMBERSHIP CERTIFICATE 

Associate members of the Society may obtain the membership 
certificate illustrated below by forwarding a request for the same to 
the General Office of the Society at 33 W. 42nd St., New York, N. Y. f 
accompanied by a remittance of one dollar. 



Society Motion Picture Engineers 




THIS IS TO CERTIFY THAT 



Society of Motion Picture Engineers 




LAPEL BUTTONS 




There is mailed to each newly elected member, upon his first 
payment of dues, a gold membership button which only members 
of the Society are entitled to wear. This button is shown twice 
actual diameter in the illustration. The letters are of gold on a 
white background. Replacements of this button may be obtained 
from the General Office of the Society at a charge of one dollar. 



SUSTAINING MEMBERS 

Agfa Ansco Corp. 
Bausch & Lomb Optical Co. 

Bell & Howell Co. 

Bell Telephone Laboratories, Inc. 

Carrier Engineering Corp. 

Case Research Laboratory 

Eastman Kodak Co. 

Electrical Research Products, Inc. 

Mole-Richardson, Inc. 

National Carbon Co. 

RCA Photophone, Inc. 

Technicolor Motion Picture Corp. 



BACK NUMBERS OF THE TRANSACTIONS AND JOURNALS 

Prior to January, 1930, the Transactions of the Society were published quar- 
terly. A limited number of these Transactions are still available and will be 
sold at the prices listed below. Those who wish to avail themselves of the op- 
portunity of acquiring these back numbers should do so quickly, as the supply 
will soon be exhausted, especially of the earlier numbers. It will be impossible 
to secure them later on as they will not be reprinted. The cost of all the available 
Transactions totals $46.25. 



1924 



No. 


Price 


1917 { I 


$0.25 
0.25 


1918 7 


0.25 


1920 


10 
11 


1.00 
1.00 


1921 


12 
13 


1.00 
1.00 


1922 


14 
15 


1.00 
1.00 


1923 


16 
17 


2.00 
2.00 



No. 


Price No. 


Price 


18 


$2.00 




29 


$1.25 


19 

20 


1.25 
1.25 


1927 


30 
31 


1.25 
1.25 


21 


1.25 




32 


1.25 


22 


1.25 




33 


2.50 


23 
24 


1.25 
1.25 


1928 


34 
35 


2.50 
2.50 


25 


1.25 




36 


2.50 


26 
27 


1.25 
1.25 


1929 


37 
38 


3.00 
3.00 


28 


1.25 







1926- 



Beginning with the January, 1930, issue, the JOURNAL of the Society has been 
issued monthly, in two volumes per year, of six issues each. Back numbers of all 
issues are available at the price of $1.50 each, a complete yearly issue totalling 
$18.00. Single copies of the current issue may be obtained for $1.50 each. 
Orders for back numbers of Transactions and JOURNALS should be placed through 
the General Office of the Society, 33 West 42nd Street, New York, N. Y., and 
should be accompanied by check or money-order. 
276 



JOURNAL 

OF THE SOCIETY OF 

MOTION PICTURE ENGINEERS 

SYLVAN HARRIS, EDITOR 
Volume XVIII MARCH, 1932 Number 3 



CONTENTS 

Page 

Two Special Sensitometers D. R. WHITE 279 

The Decibel in the Motion Picture Industry V. C. HALL 292 

Optical Instruments and Their Application in the Motion Pic- 
ture Industry I. L. NIXON 304 

Photographic Sensitometry, Part IV LOYD A. JONES 324 

Stroboscopic and Slow-Motion Moving Pictures by Means of 

Intermittent Light H. E. EDGERTON 356 

Sound in the Los Angeles Theater Los Angeles, Calif 

D. M. COLE 365 

The Reducing Action of Fixing Baths on the Silver Image 

H. D. RUSSELL AND J. I. CRABTREE 371 

Abstracts 398 

Patent Abstracts 403 

Officers 407 

Society Announcements 408 



JOURNAL 

OF THE SOCIETY OF 

MOTION PICTURE ENGINEERS 

SYLVAN HARRIS, EDITOR 



Published monthly at Easton, Pa., by the Society of Motion Picture Engineers. 

Publication Office, 20th & Northampton Sts., Easton, Pa. 
General and Editorial Office, 33 West 42nd St., New York, N. Y. 



Copyrighted, 1932, by the Society of Motion Picture Engineers, Inc. 



Subscription to non-members, $12.00 per annum; to members, $9.00 per annum, 
included in their annual membership dues; single copies, $1.50. A discount 
on subscriptions or single copies of 15 per cent is allowed to accredited agencies. 
Order from the Society of Motion Picture Engineers, Inc., 20th and Northampton 
Sts., Easton, Pa., or 33 W. 42nd St., New York, N. Y. 

Papers appearing in this Journal may be reprinted, abstracted, or abridged 
provided credit is given to the Journal of the Society of Motion Picture Engineers 
and to the author, or authors, of the papers in question. 

The Society is not responsible for statements made by authors. 

Entered as second class matter January 15, 1930, at the Post Office at Easton, 
Pa., under the Act of March 3, 1879. 



TWO SPECIAL SENSITOMETERS* 
D. R. WHITE** 



Summary. Design features of two sensitometers are presented. One of the 
sensitometers is used to make sensitometric tests on positive film under print- 
ing conditions. The other is designed to produce exposures under sound recording 
conditions. The results of tests with both of these sensitometers emphasize the im- 
portance of making sensitometric tests parallel conditions of film use, and show some 
of the errors that occur in judging speed and contrast from sensitometric data obtained 
under conditions not corresponding to the actual conditions of use. 

Sensitometric workers have found it desirable to test film under 
the conditions of its use. The time scale sensitometers frequently 
used are not representative of printing conditions where the positive 
is always exposed through a negative, nor are they ordinarily arranged 
to correspond to the conditions of sound recording where the light 
comes to the film through an optical system and has higher intensity 
and shorter exposure time than is readily obtained with the usual 
form of sector wheel. The two sensitometers herein described were 
designed, therefore, to obtain accurate sensitometric information 
under actual conditions of printing and sound recording. 

H & D PRINTER 

This machine makes use of a negative, for instance, an exposed and 
developed H & D sensitometric strip, to produce a series of graded 
exposures for testing a positive material. In this test, then, the ex- 
posure is a photographic printing operation such that the results may 
be plotted as a characteristic curve for the material tested. Provision 
is also made in this machine for comparison with pictures exposed 
and developed under similar conditions. 

A schematic view of the mechanism of the machine is shown in 
Fig. 1. The exposure timing shutter is driven by a synchronous 
motor through a pinion on the motor shaft which meshes with a 
large ring gear mounted directly on the rim of the shutter disk. Two 

* Presented in the Symposium on Laboratory Practices at the Spring, 1931, 
Meeting at Hollywood, Calif. 

** Du Pont Film Manufacturing Co., Parlin, N. J. 

279 



280 



D. R. WHITE 



[J. S. M. P. E. 



sectors are cut from this disk, one having radial sides, 30 degrees 
apart, and the other having one straight side and one side stepped 
as shown, such that each step has ten per cent greater angular opening 
than the preceding, these angles being so adjusted that the center 
step of the series has a 30-degree opening. Geared to this exposing 
shutter through a countershaft, with a total reduction of 4 to 1, is an 
auxiliary shutter which has only one sector cut from it. Two light 
houses and lights, two exposing gates, and two hand shutters are pro- 
vided, forming two independent exposing systems in which the ex- 
posure time is controlled by the one motor and shutter disk mecha- 
nism. The hand shutters in each system are not intended to control 



Ltghf 





^Exposing. 
Jhutter 



Step Test Hand 
Shutter and 
i Exposing Gate 

Hi-2) Print Hand 
Shutter and 
/Exposing Gate 




Shutter- 



FIG. 1. Schematic view of H & D printer. 

the exposure time at all but are used to prevent undesired double 
exposures. The positions of the lights are adjustable by moving the 
entire lamp support along rails provided in the light house. This 
motion allows a satisfactory inverse square law intensity variation of 
8 to 1 in the machine as constructed. The lamp supports can be 
removed from the housings intact, and placed on a bar photometer 
for color temperature and intensity measurements. 

This entire mechanism operates in such a manner that only ex- 
posures from the 30-degree exposing shutter opening for sensitometric 
tests are produced at one gate, and only exposures from the stepped 
sector for picture printing are produced at the other. With the motor 
speed and gear ratio used, the 30-degree shutter openings give expo- 



Mar., 1932] 



Two SPECIAL SENSITOMETERS 



281 



sures of one-eighteenth of a second, which is representative of printing 
exposures. 

A photograph of the machine is shown in Fig. 2. The two exposing 
gates are visible in the foreground. The light houses are about five 
feet long and extend to the rear from the compartment which houses 
the shutter disks. The countershaft and gears are also back of this 
shutter compartment and between the two light houses. One of the 
hand shutters and the signal light used as a guide in its operation ap- 
pear on the side of the machine. 




FIG. 2. Photograph of H & D printer showing the two exposing gates. 

The lamp support, removed from its housing and viewed from the 
rear, is shown in Fig. 3. With the adjustments provided, the lamp 
chosen for use may be centered, rotated on its own axis, or about an 
axis perpendicular thereto. These motions enable the selection of the 
lamp position which gives the best uniformity of intensity over the 
printing area of each exposing system. 

These design features were adopted to permit making sensitometric 
tests and picture tests under similar carefully controlled conditions. 
The sensitometric tests are made by exposures through a negative, 



282 



D. R. WHITE 



[J. S. M. P. E. 



and are so arranged that curves showing both negative and positive 
characteristics, as they are effective in the printing operation, are ob- 
tained as a result. It is not necessary to know either the positive 
or negative characteristic from other tests, for both are obtained from 
these exposures and may be compared with the photographic char- 
acteristics of the materials as shown by other methods of testing. 
The negative used in this H & D print operation may conveniently 
be in the form of an exposed and developed sensitometric strip. 
Those used in the tests here discussed were the result of time scale 







FIG. 3. 



Lamp support used in H & D 
printer. 



sector wheel exposures, with exposure times increasing by factor two 
from area to area of the negative. 

Consider, now, two exposures made through such a negative on 
this machine, in which the intensity of light incident on the negative 
is twice as great for the second exposure as for the first. The strips 
thus exposed are developed together and the resulting densities, read 
photometrically, are plotted against the log E values for the original 
negative exposure. Curves drawn through these points are shown 
in Fig. 4, where the curves labeled L and H represent the first and 
second exposures, respectively. Such curves, called reproduction 



Mar., 1932] 



Two SPECIAL SENSITOMETERS 



283 



curves, represent the relation between the density of the positive and 
the log E value for the negative. Theoretically perfect tone repro- 
duction would require these to be straight lines with slope minus one. 
Practically, this condition is neither attained nor desired, since more 
pleasing and satisfactory pictures are produced otherwise. How- 
ever, this point of tone reproduction, theory and practice, is outside 
the scope of the present paper and is only mentioned, since these 




/.o 



as- 



.J .6 .7/2 AT /.3 2.1 

FIG. 4. Two reproduction curves. Points A and B have received 
equal exposure. Point C has received one-half the exposure of B. 

reproduction curves are the first and most direct result of the opera- 
tion of this H & D printer. 

The two exposures represented by L and H had equal exposure 
times, hence it follows that any two points of equal density on the 
two curves, such as points A and B of the figure, received effectively 
equal intensities during the printing exposure. However, these 
equal printing intensities were produced under different conditions. 
For point A the effective printing intensity /^ is given by 



I A = IL X T A 



284 D. R. WHITE [j. s. M. P. E. 

where IL is the intensity incident on the negative during the ex- 
posure represented by curve L, and T A is the transmission of the 
negative at a point corresponding to A . Similarly, for the point B, 
the effective printing intensity I B is given by 

IB = IH X TB 

where I H is the intensity incident on the negative during the second 
exposure, and T B is the transmission of the negative at the point 
corresponding to B. Since, as has already been pointed out for 
these points A and B, 

I A = IB 
Therefore 

IL X TA = IH X TB 
or 

T L = TB = 2 

as 

IH 

was originally made 2 

But the difference in the effective printing densities of the negative 
between B and A, &D BA is given by 

&DBA = log ^ = log 2 = 0.3 

1 B 

It is also apparent from the method of exposure and plotting used 
for these two curves that any two points such as B and C of the figure, 
occurring at the same value on the logE scale, have densities produced 
by exposure intensities differing by factor two. This may be shown 
by the fact that, using notation similar to that above, 

IB = IH X TB 
Ic = IL X TB 

or 



T B occurs in both of the first two equations since it is the same point 
in the negative that is effective for both points B and C. 

Using these facts in the manner shown in Fig. 5, a 'series of points 
having the coordinates (D , log E ), (A, log Ej), (D z , log 2 ), . . . 
(D n , log E n ) may be found, in which the series of D's, D Q to D n , 
are the densities produced on the positive by printing exposures de- 



Mar., 1932] 



Two SPECIAL SENSITOMETERS 



285 



creasing by factor two in intensity from density to density. These 
density values may be plotted, then, at uniform spaces, log (Ep OS .) 




FIG. 5. Reproduction curves, and positive and negative characteristic 
curves resulting therefrom. 

differences of 0.3, to produce a positive characteristic curve represent- 
ing actual printing conditions. This is shown as the curve marked 




FIG. 6. Four reproduction curves may be used to increase the precision of 

the results. 



"positive" in the figure. The characteristic curve for the negative 
may be obtained by plotting at the series of values log EQ to log E n 



888 



D. R. WHITE 



Q. a M. P. E. 



on the log (E^) scale, density values increasing by uniform steps 
of 0.3 from one log value to the next, as shown in the curve marked 
"negative" in the figure. The effective printing density for any 
other log E value can be determined by interpolation from this curve. 

To improve the precision of these data, it has been more satis- 
factory to use four reproduction curves, as shown in Fig. 6, in deter- 
mining these characteristics. The results so obtained are less liable 
to error, since with the greater amount of data no single point is as 
important. 

The negative and positive characteristic curves corresponding to 



flegrfne Ounxferistic 

Visual Diffuse Density 
Effect iv* Pinnhng 

Density 



Densrty 




LogE 



O .J .6 .9 J2 tf tB 2i 

FIG. 7. Comparison of negative characteristics as obtained 
visually and by H & 



printing conditions have thus been determined without recourse to 
photometric readings of negative densities as intermediate data. 

The effective printing densities of the negatives tested to date have 
always been very close to their diffuse light densities as read on a 
photometer. Fig. 7 shows the characteristic curve of a negative as 
determined by diffuse density photometric readings and as obtained 
by this printing method. In both cases the fog reading is effectively 
subtracted by the methods used. No large density difference is 
shown here for this negative, which was developed in a metol-borax 
developer. 

The sensi tome trie characteristics of certain positive emulsions as 
determined by the procedure just outlined have been found to be dif- 



Mar., 1932] 



Two SPECIAL SENSITOMETERS 



287 



ferent from those shown by tests with the more common form of 
sensitometer, the time scale sector wheel. No general conversion 



JO 



2S 



20 



to 



O.S 



De-nsi ty 




Ttelattve Log I 



Sector Wheel 
Exposures 




,J .6 .? 12 IS / 3.1 5-t 3.7 2.0 23 26 21 12 /f 78 01 

FIG. 8. Curves showing sensitometric characteristics of a positive emul- 
sion as determined under different conditions of exposure. This emulsion 
obeys the reciprocity law within the range of the test. 



JO 



2S 



20 



I.S 



10 



OS 



HtD Printer 
Exposures 



Density 



Relative Log I 



Sector Wheel 
Exposures 




3 .6 .9 1.2 I.S 18 S, j.4 37 2.0 5.J 26 21 7.2 U 78 ./ 

FIG. 9. Curves showing sensitometric characteristics of a positive emul- 
sion as determined under different conditions of exposure. This emulsion 
shows reciprocity law failure. 



factor from one type of exposure to another is possible, since not all 
emulsions show the same variation in characteristic. 

Fig. 8 shows the characteristics of one positive emulsion as shown 



288 D. R. WHITE [J. S. M. P. E. 

by a group of tests. The six curves under the heading "sector wheel 
exposures" show time scale characteristic curves determined at six 
exposure intensity levels, increasing right to left by factor two from 
curve to curve in the group. The printing characteristic is shown by 
itself, as determined by the H & D printer. No failure of the re- 
ciprocity law is observed in the range covered by these data. The 
contrast of the material appears independent of the exposure, and 
therefore only dependent on the development which the film received. 
However, this condition of constancy of contrast does not hold 
for the positive emulsion represented by Fig. 9. This emulsion was 
prepared under different conditions from the first. Here, there is a 
noticeable reciprocity law failure even within the range of time 
intensity variations shown in the sector wheel tests. Gamma de- 
pends on both the time and intensity used in its determination. The 
contrast values shown by the sector wheel curves made with the higher 




Objective 

V_y Condenser 

Hand Shutter 



u 



Ribbon nhmenf Lam/3 

FIG. 10. Schematic diagram of the optical system of the high 
intensity sensitometer. 

exposure levels become closely equal to that shown by the H & D 
printer, but differ appreciably from the values obtained at lower 
intensities. 

HIGH INTENSITY SENSITOMETER 

The second of the two special machines here discussed was designed 
to test film under conditions corresponding to those of sound record- 
ing. 

The optical system of this machine is shown schematically in Fig. 
10. A filament image from a ribbon filament light is formed at the 
slit by a condenser lens. The slit in turn is imaged on the photo- 
graphic film carried on the film drum by a microscope objective. A 
series of fixed slits of various widths is provided to obtain the exposure 
range desired. This construction was used in preference to any form 
of calibrated light valve as its constancy from time to time is assured 






Mar., 1932] 



Two SPECIAL SENSITOMETERS 



without difficulty. This group of slits is mounted on a circular plate 
driven from the film drum shaft through an intermittent motion and 
gearing. The final accurate positioning of the slits in the correct 
position in the optical system is assured during operation of the 
machine by a wedge fitting a V-groove in the rim of the slit carrying 
plate. The wedge is disengaged by a cam when the intermittent 
motion is about to operate and is re-engaged before exposure. The 




FIG. 11. 



Photograph of the high intensity sensitometer showing the film 
drum, slit carrying plate, and optical system. 



mechanism is adjusted to make a complete series of eleven exposures 
on a single turn of film fastened to the film drum. The hand shutter 
placed in the optical system is used to prevent double exposure and 
to permit the machine to come up to speed, 90 feet per minute of the 
film, before any exposures are made. 

Fig. 11 presents a view of the machine showing particularly the 
film drum, slit plate, and the optical system. The disk with the small 
sector cut from it is an auxiliary shutter which limits the exposures 



290 



D. R. WHITE 



[J. S. M. P. E. 



to the desired portions of the film being exposed and which cuts off 
the exposure during the motion of the slit carrying plate. 

Fig. 12 is taken from a different viewpoint and shows more clearly 
the intermittent motion and slit positioning mechanism. 

In making the machine, difficulty was anticipated and experienced 
in making the slits to exact spacings. Rather than spend too great 
an amount of time on an unnecessary detail, the slits were set at 
approximately the values desired and then finally calibrated in place 




FIG. 12. A view of the high intensity sensitometer showing some of the 

mechanism. 



by measuring their light transmission with a photoelectric cell. 
These slits vary from approximately 0.00025 to 0.008 inch. With 
eleven slits the average step is about factor \/2, but the steps are 
not entirely uniform. 

A comparison of the results of tests with this machine and with a 
time scale sector wheel with much lower intensity and longer expo- 
sures is shown in Table I. In this comparison the development was 
carried to a point representing variable density recording condi- 
tions. Compared emulsion by emulsion, the gammas produced by 



Mar., 1932] TWO SPECIAL SENSITOMETERS 291 

the two methods of exposure are not widely different, though with 
two emulsions the sector wheel gamma is higher. In the relative 
emulsion speeds, as determined by each experimental method, larger 
differences are found. In the table, emulsion A is arbitrarily taken 
as having a speed of unity in both tests, and the others are evaluated 
in comparison with it. 

TABLE I 

Relative Speed and Contrast Values as Obtained by Exposures with a Sector Wheel 
and with the High Intensity Sensitometer 

Gamma Relative Speed 



Emulsion 


Type 


Developed 


H.I. 


S.w. 


H.I. 


S.W. 


A 


Positive 


2 1 /z min. 


0.50 


0.49 


1.0 


1.0 


B 


Positive 


2 l /z min. 


0.55 


0.64 


1.7 


1.3 


C 


Positive 


2*/2 min. 


0.53 


0.51 


1.1 


1.0 


D 


Fast pos. 


2 l /z min. 


0.58 


0.66 


2.6 


2.2 


E 


Sp. recording 


3 min. 


0.49 


0.50 


3.0 


2.2 


F 


Sp. recording 


3 min. 


0.52 


0.54 


3.5 


2.7 



In obtaining these data, individual development characteristics 
were ignored, but a difference was introduced to compensate for the 
lower gamma infinity of two of the special sound recording emulsions. 
The residual variations in gamma prevent an accurate statement 
of the relative emulsion speeds, but the possible errors due to this 
cause are much less than the differences found. Thus, it is evident 
that tests made under the one set of conditions furnish no sure guide 
to emulsion characteristics effective under the other set of conditions. 
The results of all these tests on both types of sensitometers emphasize 
again the necessity for care in selecting test conditions in photo- 
graphic work. The two machines described have worked out quite 
satisfactorily, each meeting the special needs in its own field. 



ERRATUM 

The following corrections should be made in the paper, A Method of Measuring 
Directly the Distortion in Audio Frequency Amplifier Systems, by W. N. Tuttle, 
beginning on page 199 of the February, 1932, issue of the JOURNAL: 

A square root sign should be placed over the entire numerator of the right- 
hand member of the equation on page 200. On page 205, line 2, the symbols 
"2-C" should read "a-c." 



THE DECIBEL IN THE MOTION PICTURE INDUSTRY* 

V. C. HALL** 

Summary. The development of the term "decibel" is outlined, and its convenience 
in measuring the characteristics of electrical circuits discussed. The relation of 
the decibel to photographic density is pointed out and illustrated by calculations of 
the effect of ground noise reduction devices. Finally, the values of acoustic power 
of common sources of sound are given as the levels in decibels based on various reference 
points in use in the motion picture industry. 

The use of the term "decibel" has increased rapidly since its 
introduction into the motion picture industry because it is a con- 
venient method of handling quantities which might otherwise lead 
to cumbersome expressions. Many publications have been written 
concerning the decibel and its historical development and its applica- 
tion to the problems under consideration. These papers have been 
freely drawn from for examples and data for this summary. No 
new material is presented but it is hoped that a review of some of 
the ways in which the decibel enters into sound motion pictures 
may prove of value to those who are not familiar with this unit. 

Development of the Term. The decibel (db.) is the name which 
was chosen for the transmission unit (TU), in terms of which a 
great deal of telephone and talking motion picture apparatus is 
calibrated. Their values are identical, and the name itself was 
suggested several years before its final adoption. Units of similar 
type had been universal in telegraphy and telephony for many 
years, and came into being from the fact that an electric current 
representing a certain amount of power loses a certain fraction of 
that power for every mile, let us say, of transmission line over which 
it travels. The unit of electric power, representing the rate of 
doing electric work, is the familiar watt, in terms of which most 
electrical equipment is specified. If, then, a power of ten watts 
is started out over a telegraph line, it would drop to 0.9 of this in 
perhaps two miles. This nine watts would drop to 0.9 of 9, or 0.81, 

* Presented at the Spring, 1931, Meeting at Hollywood, Calif. 
** Research Laboratories, Eastman Kodak Co., Rochester, N. Y. 
292 



DECIBEL IN MOTION PICTURES 293 

in the next two miles, and so on, the power decreasing the same 
fraction every two miles, independently of the actual amount of 
power starting over this particular section of line. The multiplying 
together of these factors was laborious, so the practice grew up of 
expressing a loss of power in terms of the number of miles of telephone 
line which would decrease the power by the same amount. Thus, 
if the power dropped to 0.81 in one section of a circuit, and to 0.73 
in another, the total loss could be found by multiplying 0.81 by 
0.73, which equals 0.591. It is much simpler, however, to say that 
the first section has a loss of two miles and the second a loss of 
three miles, and that the two together have a loss of two plus three 
or five miles. The substitution of addition of numbers for the 
multiplication of corresponding numbers is a property of logarithms, 
and a system which allows this to be done is called a logarithmic 
system. Thus the use of the "mile of standard cable" constituted 
a logarithmic system, standards for which were set up by those 
doing telephone work. It was natural that the properties of cables 
should change as improvements were made so that eventually the 
old standard became inconvenient. Also, amplifiers were developed 
in which the output power was more than the input power so that 
the circuit had a "gain" instead of a "loss." These developments 
led to the adoption of a simpler unit, based only on the relation 
between the power output of the circuit and the power input. The 
expression used was 

/watts output \ 

Number of telephone transmission units = logarithm ( : I 

V watts input / 

The values obtained from the above are positive when the output 
is greater than the input, and negative when the output is less than 
the input, indicating a loss. This unit as defined above happened 
to be a rather large one considering the ratios of power encountered 
so that the unit actually adopted was arbitrarily chosen to be 0.1 
of this. Accordingly, in practice the equation becomes 

Number of transmission units or TU's = 10 X logarithm 

The fundamental unit was named the "bel," after Alexander 
Graham Bell, with the spelling simplified to avoid confusion. Since 
the name is for the larger unit, the prefix "deci" was affixed to the 
name to indicate its derivation, and now we have: 



294 V. C. HALL [J. S. M. p. E. 

Number of transmission units or TU's = number of decibels or number of 

/watts output \ 
db ' = 10Xloganthm ( watts input) 

Method of Calculation. In order to get some idea of the magnitude 
of the decibel, the following short table is presented, giving a list 
of power ratios from 1 to 10, with the corresponding number of 
decibels. These are found simply by looking up the logarithm of 
the ratio in a table of common logarithms, or on a slide rule, and 
multiplying by 10. 

TABLE I 
Power Ratio Decibels Power Ratio Decibels 

1 0.0 1.0 -0.0 

2 3.0 0.5 -3.0 

3 4.8 0.33 -4.8 

4 6.0 0.25 -6.0 

5 7.0 0.20 -7.0 

6 7.8 0.17 -7.8 

7 8.4 0.14 -8.4 

8 9.0 0.125 -9.0 

9 9.6 0.111 -9.6 
10 10.0 0.100 -10.0 

In the first two columns of Table I the output power is assumed 
to be greater than that at the input or, in other words, an ampli- 
fication instead of a loss is assumed. When this is true, the output 
is said to be so many decibels above the input. When a loss occurs, 
as in the last two columns, the input is of course above the output, 
and the output is said to be down so many decibels. It will be 
noted that the power ratios in the second case are the reciprocals of 
the corresponding ratios in the first case, while the number of decibels 
is the same, with a minus sign. This means, for example, that 

10 x logf = -10 X log^ 


which in turn follows from the general principle in logarithms that 
the logarithm of a number is equal to minus the logarithm of its 
reciprocal, which is another way of saying, minus the logarithm of 
one divided by the number. 

If the amplifications of a group of separate amplifying units are, 
respectively, 3, 7, 10, and 4, the total amplification, if they are 
connected in series, can be found by multiplying these numbers 
together. This gives 3 X 7 X 10 X 4 = 840 times If the ampli- 
fications are expressed in decibels, however, this becomes simply 



Mar., 1932] DECIBEL IN MOTION PICTURES 295 

4.8 + 8.4 + 10 + 6.0 = 28.2 decibels. In this case the result is 
about as simple to calculate in one way as the other. If instead of 
referring to amplifications, however, the numbers referred to losses, 
and the power were reduced to 0.33, 0.14, 0.1, and 0.25 of the original 
value, respectively, the multiplication of factors would be more 
difficult. Looking up the corresponding decibels in the second part 
of Table I gives -4.8, -8.4, -10, and -6.0, respectively, and 
the sum of these equals 28.2 decibels, which is the total power 
loss of the group. In practice it is rare to find the even power ratios, 
while the decibel equivalents are usually expressed to sufficient 
accuracy by two or, at the most, three figures, the addition of which 
is obviously more quickly done than the corresponding multiplica- 
tion, while the chance of error is also greatly reduced. 

From Table I other values of either decibels or power ratios can 
easily be found. For instance, 26 decibels is made up of 10 + 10 + 6, 
corresponding to power ratios of 10, 10, and 4, respectively, and the 
product of these, 400, is the power ratio corresponding to 26 decibels. 
Similarly, a power ratio of 75 can be divided into its factors 3 X 5 X 
5, and we have 4.8 + 7.0 -f 7.0 = 18.8 decibels, which corresponds 
to a power ratio of 75. 

The decibel is a convenient unit as the ear can just recognize a 
change in the volume of sound corresponding to an attenuation or 
gain of one decibel, equivalent to about 12 per cent. For this 
reason, volume controls and faders calibrated in decibels give a very 
uniform increase in the volume as the ear recognizes the changes as 
equal steps. Although the ear can detect a change of one decibel, 
most volume controls are set to change the volume by about 3 db., as 
this step is not so great that a change of one step will raise the volume 
from too low to too high, and the reduction of the number of steps 
simplifies the apparatus. For the control of volume in sound re- 
cording where a smooth stepless change is wanted, a potentiometer 
or volume control similar to that on a radio receiver is used, and the 
scale is graduated in decibels for certain intervals on the dial. 

Methods of Measurement. So far there has been no mention of 
the method by which the measurements are made. In the electric 
circuits used in talking motion pictures it is usually too difficult 
to measure the power, in watts, directly, since the frequency of the 
alternating currents which must be measured varies all the way 
from 30 to 10,000 cycles per second, and the power levels are of the 
order of thousandths of watts. Both of these factors tend to make 



296 V. C. HALL [j. s. M. p. E. 

the use of wattmeters impracticable, so most results are obtained 
by measuring either the current through a known resistance, or the 
voltage across a known resistance. If it is necessary to know the 
power it may be found from the following relations which exist 
between the various electric quantities. 

Power (watts) = current X voltage (amperes X volts) 
Voltage (volts) = current (amperes) X impedance (ohms) 

From these two are derived : 

Power (watts) = (current) 2 X impedance 
= (voltage) 2 /impedance 

In order further to simplify the circuits electrically, most sound 
motion picture apparatus, and a great deal of telephone equipment, 
is designed so that the impedances in both the input and output are 
equal, the actual value usually being approximately 500 ohms. 
This is true of amplifiers, niters, equalizers, volume controls, etc., 
and further simplifies the calculation of losses or gains in the various 
units as follows: If we substitute for the power input and output 
in the formula for the decibel the expression for the power in terms 
of the current and resistance, it becomes: 

. ,. (current output) 2 X impedance 

Number of db. = 10 X log ; 

(current input) 2 X impedance 

and the resistance cancels out, leaving the formula 

. ,. (current output) 2 (A) 2 

Number of db. = 10 X log -. . ' ; = 10 X log ^f, 

(current input) 2 6 (7 2 ) 2 

letting 7 2 stand for the current input, and /i for the current output. 
From the principle that the logarithm of the square of a number is 
equal to twice the log of the number it is possible to write, instead 
of the relationship above, 

Number of decibels = 10 X 2 X log ^ = 20 log 

U2J (1-1) 

It is important to note that while the power is independent of 
the impedance, the above formula is true only when the impedances 
through which the currents /i and 7 2 are flowing are equal. If it 
is more convenient to measure the voltage across a known impedance 
than the current through it, as is often the case, we can write, again 
assuming that the impedances are equal in the two places the mea- 
surements are made, 

Number of db. = 10 log ( . VO ' tage "tput)Vi mpe dance 
(voltage mput) 2 /impedance 



Mar., 1932] DECIBEL IN MOTION PICTURES 297 

The impedance cancels out as before, and the expression becomes 
Number of db. = 10 log = 20 l S 



which is exactly the same as if the current output and input were 
measured. The most usual methods of measuring electric currents 
of audio frequency as found in sound motion picture work are by 
the hot wire ammeter and the vacuum thermojunction. The 
thermojunction is the more accurate, and in it the current heats a 
length of resistance wire in a vacuum, while near the center of the 
wire is fastened a junction of two wires of different composition. 
When this junction is heated a voltage is developed which depends 
on the nature of the two metals and the temperature to which it is 

TABLE II 

Voltage or 

Power Ratio Current Ratio Decibels 

1 1.0 0.0 

2 1.4 3.0 

3 1.7 4.8 

4 2.0 6.0 

5 2.2 7.0 

6 2.45 7.8 

7 2.64 8.4 

8 2.83 9.0 

9 3.00 9.6 
10 3.16 10.0 
20 4.47 13.0 
40 6.33 16.0 

100 10.0 20.0 

heated. A meter connected across the other ends of the two wires 
will deflect in accordance with the current variations, and if cali- 
brated in terms of a standard meter can be used to measure the 
current accurately. The volume indicator, which is essentially of 
the vacuum tube voltmeter type, although it may draw a small 
current, due to its ruggedness is the most popular of audio frequency 
measuring instruments. In this the voltage across a resistance 
changes the grid voltage on a vacuum tube which is so connected 
that the change causes a variation in the steady plate current of 
the tube. A milliammeter in the plate circuit then reads in ac- 
cordance with the changes in voltage across the resistance. Most 
volume indicators are built on this principle although, since the 
entire scale of the meter corresponds to a relatively small number 



298 V. C. HALL [J. s. M. p. E. 

of db., most of them have a resistance device connected so that the 
readings may be cut down by fixed amounts, say, 2 db. per step. 
By using this device the needle of the meter can be kept at one 
position on the scale as the voltage varies, and the db. change noted 
by the change in the setting of the control mechanism. 

In Table II are shown the power ratios, voltage or current ratios, 
and the decibel changes which correspond to them. Thus, if the 
measured current ratios between output and input of three amplifiers 
are 2.45, 4.5, and 6.3, the total gain is 

7.8 
13.0 
16.0 

36.8db. 

If the composite voltage or current gain is desired it is found by 
36.8 = 20 log (voltage ratio) 

from which log (voltage ratio) = 36.8/20 = 1.84, and voltage 
ratio = 69.1 

Reference Levels for Calibration of Apparatus. While from its 
definition the difference in decibels between any two amounts of 
power is calculated without reference to any standard power, it is 
convenient to have some value of power which may be considered 
as a reference level. Power is always expressed in watts, so that 
at first thought it might seem obvious that one watt of power would 
be the correct unit to choose. The amount of power encountered 
in either electrical or acoustic measurements in sound motion pictures, 
however, is nearly always much smaller than one watt, and the 
expression of levels when the ratio of power is less than unity is 
negative. This unit would involve the use of negative numbers 
for the expression of nearly all powers measured, and would prove 
inconvenient. 

In acoustic measurements, the power of a sound wave in air is 
usually very small and is spread throughout a considerable volume 
of space. To simplify calculations it is generally assumed that the 
sound waves radiate uniformly in a hemisphere from the source. 
Results indicate that the assumption is justified, provided that a 
reasonable distance from the source is allowed and that no difficulty 
is encountered from reflections from walls, ceilings, etc. 



Mar., 1932] DECIBEL IN MOTION PICTURES 299 

The energy in the sound wave may be measured either in its 
entirety, or as the amount passing through a unit area (usually 
a square centimeter) at certain distances from the source. Since 
the smallest amount of energy which can excite the sensation 
of hearing is the smallest amount of useful energy a sound wave can 
have, this value is taken as the "audibility threshold" or "acoustic 
level," and is usually considered to be about 4 X 10~ 16 watts per 
square centimeter. 

Another value sometimes used is the "phonic level," which is 
simply one microwatt per square centimeter of cross-section of the 
air through which the wave is traveling. 

In the electric circuits associated with sound motion pictures, 
the powers vary in value from as low as the audibility threshold 
up to several watts, as the power necessary to operate loud speakers 
in theaters satisfactorily may in extreme cases be as great as 15 
watts or more. The general levels at which measurements can be 
made easily correspond to a few milliwatts, and volume indicators 
are usually calibrated to read "0" level at about 6 milliwatts. 

Relation between the Decibel and Photographic Density. The 
amount by which the silver deposit on a photographic film reduces 
the amount of light transmitted by the film is expressed by a logarith- 
mic unit called density. The first measurements of the decrease of 
light intensity were made by observing the percentage of the incident 
light which the film transmitted. Various considerations led to 
the adoption of a logarithmic unit which is defined as the logarithm 
of the reciprocal of the fractional transmission, thus density = 
log 1/r, where T is the fraction of the light transmitted by the silver 
deposit. Since the value of the light transmitted is always less 
than that which is incident, this fraction is always greater than 
unity, all photographic densities being positive, and varying in 
practice from 0.04, the ratio in which clear film base reduces the 
light transmitted to about 6.0, representing opacity for all practical 
purposes. 

When a variable density sound record passes through a projector 
the changes of density cause the light of the exciting lamp which is 
incident on the photoelectric cell to vary in intensity. This causes 
the photoelectric current to vary, but as it is proportional to the 
light striking it, the change of current is proportional to the trans- 
mission of the film, and not to the density. Thus, if a transmission 
TI corresponds to a current /i, and if the transmission should change 



300 V. C. HALL [J. S. M. P. E. 

to T 2 , the photoelectric current would change proportionally to J 2 , so 

that it can be written: 

7\ = /i 

This can be changed to log ~ = log by taking logarithms of both 

J. 2 -*2 

sides of the equation, and can be multiplied by 20 also without chang- 
ing the validity of the statement. This leaves 

20 log ~ = 20 log 

2-2 12 

The left-hand side of this equation is identical in form with the ex- 
pression for the power reckoned in decibels when two currents act 
through equal resistances; and since the photoelectric currents in 
this case both pass through the same resistance (the amplifier input 
resistance) we can substitute the decibel for this part of the expres- 
sion. We have 

Number of db. = 20 log -=^ 

The right-hand side contains the logarithm of the ratio of two photo- 
graphic transmissions. These can be written as the product of 
V TI X TI, and since the logarithm of a product is equal to the sum 
of the logarithms, it becomes 

Number of db. = 20 Aog ^ + log 

\ * 2 

or 

Number of db. = 20 ( loe^r - 



The logarithm of l/7\ is no more than the density corresponding 
to this transmission (Di), and log l/r 2 equals the corresponding 
density (D 2 ). We may therefore substitute these values and reach 
the expression 

Number of db. = 20 (D 2 - A) 

showing that any change in photographic density, multiplied by 
20, gives the corresponding change in electrical power in decibels. 
From these statements it is possible to calculate the efficiency of 
the noise reduction units now in use in light valve recording studios. 
The "ground noise" arising in sound motion pictures is due partly 
to slight irregularities in the current which are inherent in the photo- 
electric cell, but chiefly to the changes in current caused by dirt 



Mar., 1932] DECIBEL IN MOTION PICTURES 301 

and scratches in the photographic film. The effect of a scratch or 
particle of dirt is to cut the light down by a certain fraction, so that 
its effect on the photoelectric current will be less in proportion to 
the amount of light which is left to be affected. Therefore, whatever 
can be added to the average density of the positive sound track will 
help to reduce both types of ground noise in an amount equal in 
decibels to 20 times the difference between the two densities. It 
must be noted that an increase in the average density of an ordinary 
sound print does not cut down this noise, as the volume it is possible 
to get also is cut down, so that the amplification must be raised, 
restoring not only the signal, but also the noise, to its former level. 
It is only by cutting down the light while the sound volume is low, 
or during silent passages, that any effect is found, and if the density 
of the film can be decreased to its normal value when the sound 
volume increases, the amplification does not have to be increased 
to keep the proper level in the theater. The amount by which 
the noise may be decreased depends fundamentally on the amount 
by which the valve may be closed in recording. This narrowing of 
the light valve slit is accomplished by sending a direct current 
through the valve in addition to the amplified signal coming from 
the microphone. This decreases the density of the negative during 
sound passages of low volume, increasing the density of the positive 
during the same sequences. In following through the theory of 
sound recording and reproducing by the light valve method, proper 
sound reproduction depends on the proportionality of the movement 
of the light valve strings to the sound pressure at the microphone 
in recording, and the photographic processing must be such that 
the transmission of the positive is also proportional to the sound 
pressure. Therefore the transmission of the positive must be pro- 
portional to the valve opening. If the normal slit width is 1.0 
mil (0.001 inch), and it is biased in noiseless recording to 0.3 mil, 
the positive will have a transmission TI for the 1.0 mil slit and a 
transmission T z for the 0.3 mil slit. The proportionality equation is 

1.0/0.3 = 7yr 2 

taking logarithms of both sides 

log 3.3 = log 7yr 2 . 

It has been shown that log Ti/T z = (D z A), so the above becomes 
DZ DI = 0.52, and since the reduction in noise is 20 times the 



302 V. C. HALL [J. S. M. P. E. 

change in density, in this case it becomes 10.4 decibels. As has 
been stated this would be about three steps on an ordinary fader 
and would be very appreciable. 

In the variable width method of recording, during silent passages 
one-half the sound track has a high density and one-half is clear. 
In order to reduce the noise due to transmission of light through the 
clear area, a mask is arranged in making the negative to cut off the 
light incident to this area. During printing this area is printed 
to a high density, leaving only a very narrow unexposed line in the 
center of the record. As the sound intensity increases, the mask 
is moved farther and farther over, leaving more and more of the 
sound track available for the making of the record. In such a case, 
the intensity of noise is again dependent on the amount of light 
transmitted, but since the film is either clear, letting through all the 
light, or so dense as to allow practically none, the amount of light 
transmitted is proportional to the width of the clear portion of the 
track. The decrease in the noise intensity, following the same line 
of reasoning as before, will depend on the width of track it is neces- 
sary to leave in the center during silent passages. This is at least 
5 mils, and since in the variable width recording system the width 
of the whole sound track is 70 mils, the clear portion is normally 
35 mils. The reduction in the intensity of the noise can be cal- 
culated from the change in the photoelectric current, which depends 
on the width of the clear track. Thus from equation 

Number of db. = 20 log 35/5 = 20 log 7 = 20 X 0.845 
Number of db. = 16.9 or approximately 17 db. 

Conclusion. It is hoped that the foregoing explanation of the 
various ways in which the decibel enters into the sound motion 
picture may prove of value to those who find that the literature of 
the art includes many statements which depend, for a complete 
understanding, on an accurate conception of exactly what the 
function of the unit is, and the reasons why its use is convenient. 
For reference a tabulation of the various levels occurring in parts of 
sound motion picture systems is added. These levels are in some 
cases only approximate, owing to their nature, but they indicate 
the order of magnitude to be expected. 

Distances as given refer to columns A, B, and C. Radiation is 
assumed to be in the form of a hemisphere with the power given in the 
first column generated at the center. (A), decibels above audibility 



Mar., 1932] DECIBEL IN MOTION PICTURES 303 

threshold (acoustic level, assuming 4 X 10 ~ 10 microwatts per sq. 
cm.); (B), decibels above one microwatt per sq. cm. (phonic level); 
(C), decibels above 0.006 watts (electrical level) of output of con- 
denser microphone into 25 megohms input. 

TABLE III 

Source of Sound 
Total Power Microwatts ABC Distance 

Soft whisper 0.001 17 -77 -144 3 feet 

Average speech 10 57 -37 -104 3 feet 

Very loud speech 1000 77 -17 -84 3 feet 

Peak of speech 5000 84 -10 -77 3 feet 

Peak of singing 30000 91.8 -2.2 -69.2 3 feet 

Soft violin in orchestra 4 43-51 -118 10 feet 

Piano average 4000 73 -21 - 88 10 feet 

Piano highest peak 2 X 10 6 100 +6 - 61 10 feet 

Bass drum peak 25 X 10 6 107 +13 - 54 15 feet 

75 piece orchestra peak 66 X 10 6 113 +19 -48 15 feet 

Pipe organ peak 13 X 10 6 105 +11 - 66 15 feet 

No attempt has been made to quote sources of data given in the 
course of the paper. The data for Table III were derived from the 
first two references given and further references will be found in 
the following list. 

REFERENCES 

FLETCHER, H.: "Speech and Hearing," D. Van Nostrand Co., Inc., New 
York, N. Y. (1929). 

SIVIAN, L. F., DUNN, H. K., AND WHITE, S. D.: "Amplitudes and Spectra 
of Certain Musical Instruments and Orchestras," /. Acoustical Soc. Amer., 
2 (January, 1931), p. 330. 

WOLF, S. K., AND SETTE, W. J.: "Acoustic Power Levels in Sound Re- 
production," /. Acoustical Soc, Amer., 2 (January, 1931), p. 384. 

DREHER, C.: "Progress in Sound Picture Recording," Electronics, 2 (March, 
1931), p. 542. 

MARTIN, W. H.: "Decibel the Name for the Transmission Unit," Bell 
System Tech. Jour., 8 (January, 1929), p. 1. 

SHEA, T. E.: "Transmission Networks and Wave Filters," D. Van Nostrand 
Co., Inc., New York, N. Y. (1929), p. 43. 



OPTICAL INSTRUMENTS AND THEIR APPLICATION IN 
THE MOTION PICTURE INDUSTRY* 

I. L. NIXON** 



Summary. This paper deals not with the optics of the photographic lens, motion 
picture projector, or studio illuminator, but rather with those instruments such as 
microscopes, photometers, etc., the use of which has contributed greatly to the advance 
of the motion picture art of today. A simple explanation is given of the different 
types of instruments and the general optical principles involved, and some of their 
specific applications, which indicate the debt which industry owes to optical science. 

When speaking of optics in the motion picture industry, it is but 
natural for those of us who are most intimately connected with the 
industry, to think of optics as applied to the photographic lens as 
used on the motion picture camera, or to the optics of the projector, or 
for illumination in the studio, but I purpose to outline briefly some of 
the different types of optical instruments that have been used or 
might be used in the development of new materials, control of proc- 
esses, and control of accuracy of parts. Because of the fact that 
many of these devices have been used largely in a research way and 
not in sufficient numbers to attract attention, they might be classed 
as the modest group of silent workers that have made the high perfec- 
tion of the present art possible and that will play an important part 
in the achievements of the future. 

The microscope of one form or another is probably the most widely 
used optical instrument in the motion picture industry. It hardly 
seems necessary to define a microscope, but it might be described 
generally as a device having a system of lenses, suitably supported by 
mechanical arrangements, which will produce a magnified image of a 
small object so that the eye may distinguish between details of 
structure not otherwise discernible. 

A simple magnifying glass might be considered as qualifying as a 
microscope under this definition, but this paper will deal with what 
may be termed a compound microscope, a typical one being repre- 

* Presented at the Spring, 1931, Meeting at Hollywood, Calif. 
** Bausch and Lomb Optical Co., Rochester, N. Y. 
304 



OPTICAL INSTRUMENTS 



305 



sented by Fig. 1, where a system of lenses, mounted together and 
known as an objective, is attached to the lower end of what is known as 
a body tube and another system known as the eyepiece is mounted 
in the upper end of the tube. The objective acts as a photographic 
lens would act, and forms a magnified image of the object in the focal 
plane of the eye lens which, in turn, magnifies that image. Hence we 
have compound magnification and in turn a compound microscope. 
By varying the power of one or both of these units the magnification 
is accordingly changed, the range of magnification being from IOX 
to approximately 2000X. 




FIG. 1. A typical compound microscope. 

The design and accuracy of the mechanical parts of such an instru- 
ment are quite as essential to its functioning as are the optical parts. 
It must be substantially constructed, and yet a certain symmetry 
of design is demanded and its movable parts must be accurately fitted 
and free from any lost motion, yet immediately responsive to adjust- 
ment. 

The mechanical part of the compound microscope is referred to as 
the stand and consists of the following general parts: 

(A} Base, of a design that will have sufficient spread and weight to assure the 
stability of the instrument in either an upright or inclined position. 



306 I. L. NIXON [j. s. M. P. E. 

(5) Stage, on which is placed the object to be observed. This may be either 
plain, rectangular, or circular revolving, and both styles may be fitted 
with mechanical devices for moving the object in two directions at 90 
degrees to each other for easy searching of the specimen. These ad- 
justments may also be provided with scales for relocation of the specimen 
if desired. 

(C) Arm, attached to the base and supporting the body tube with its adjust- 

ments. 

(D) Body tube. 
() Objective. 

(F) Eyepiece. 

(G) Coarse adjustment, by rack and pinion, which must move easily and yet 

be free from lost motion. 
(H) Fine adjustment. 
(/) The substage with condenser or illuminating lens system, which functions 

either in conjunction with daylight or a suitable artificial light source to 

illuminate the specimen efficiently, if it be one with which transmitted 

light may be used. 

In Fig. 2 we have shown diagrammatically the path of light of such 
a microscope, which seems to need no further explanation except to 
point out that when looking into the microscope the image appears as 
though it was being viewed at a point 10 inches below the equipment. 
If a screen is held 10 inches above the eyepiece an image will be 
formed at that plane equal in magnification to the image observed in 
the eyepiece and the magnification would increase proportionately 
as the distance was increased beyond 10 inches. 

This represents the typical biological or medical type of microscope, 
large numbers of which are manufactured annually for use in the 
schools and colleges, but which are being used more extensively each 
year in industrial laboratories where transmitted light may be used. 

A number of deviations from this typical instrument in the way of 
special illuminating devices and accessories of one sort or another 
make the equipment particularly suited for some specialized work. 
Before passing on to these, however, it will be interesting to note the 
similarity of the optical system of the microscope to that of the motion 
picture projector. A light source with the substage condenser cor- 
responds to the light source and the condensing lens system; the 
stage on which the specimen is placed may be compared to that of 
the film gate supporting the film, and the objective lens and eyepiece 
may be considered as one unit corresponding to the projecting lens. 

In addition to using a microscope as a device for studying the 
structure of materials it may also be used for measuring the size of 



Mar., 1932] 



OPTICAL INSTRUMENTS 



307 



particles or parts, or their separation, by the use of a ruled disk to be 
placed in the eyepiece at what is known as the diaphragm plane or in 
the same plane as that of the image formed by the objective so that 
both the scale and the image of the object will be in the focus of the 
eye lens. (Fig. 3.) 

Such a scale may be ruled with divisions to represent a definite 
value on the specimen (0.001 of an inch, for instance) for use with 
definite combination of eyepiece and objective producing a fixed 
magnification, or the eyepiece disk may be ruled in definite values, 




FIG. 2. Diagram showing path of 
light of a typical microscope. 

and a stage micrometer used to evaluate the rulings on the eyepiece 
disk, according to the combination of objective and eyepiece being 
used. The use of such a device was probably first used by the doctor 
in counting the number of blood corpuscles per given quantity of 
solution, but has been adopted by the industry as a means of mea- 
suring and determining the distribution of silver grains in emulsion. 

As evidence that the microscope is being recognized as one of the 
most important tools in modern industry is the fact that a number of 
the leading universities are introducing as a division in the chemical 
engineering courses one known as "Chemical Microscopy" in which 



308 



I. L. NIXON 



[J. S. M. P. E. 



the principles of the microscope, its applications, and the interpreta- 
tion of the results are taught. 

There is no industry that I know of in which a microscope could 





FIG. 3. Photomicrograph of sil- 
ver grains with micrometer scale. 



FIG. 4. Path of light of special 
microscope for examination of 
paper surfaces. 



not be used to advantage in the control of its raw materials and 
finished product. 

The number of ways in which a microscope is used in the laboratory 




FIG. 5. Special paper microscope. 



of a manufacturer of film is amazing. An almost constant study is 
made in the size, shape, and distribution of silver grains in the emulsion 
both before and after development. Photomicrographs are fre- 



Mar., 1932] 



OPTICAL INSTRUMENTS 



309 



quently made for record and control purposes from which frequency 
curves may be plotted if desired. 

We do not ordinarily think of the finished film being built up of a 
series of layers, but it is; and each one of the processes contributing to 
this building up must be carefully controlled. When something goes 
wrong, they send for the trouble shooter, the man with the micro- 
scope. A cross-section of the film will probably be made with a 
microtome, a device for making sections only a few microns in thick- 
ness; and when this is observed one clearly sees these layers of differ- 
ent materials, and the trouble can usually be traced to a certain opera- 
tion or to impurities that are causing the trouble. 

Standards of surface finish of both film and paper may be set up 





FIG. 6. Photomicrograph of a 
paper surface at 40 magnifications. 



FIG. 7. Path of light when il- 
luminating opaque objects. 



according to microscopical specifications and in the event of trouble it 
is fairly easy to trace back against the standard and locate the source. 

Dr. L. A. Jones, of the Eastman Kodak Co., in his study of paper 
surfaces decided that there was needed a special type of illumination 
with provision that the exact illumination could be duplicated at any 
time, because surface appearance of paper depends so much on the 
amount and angle of illumination. There was developed a micro- 
scope with an illuminating system as indicated in Fig. 4. 

A light source and condensing system produces a beam of light 
passing through a ground glass at G, which strikes the 45-degree 
annular reflector M , and the light is reflected upon the specimen at 0. 
It is obvious that this is annular illumination which illuminates equally 
from all directions, and with no direct top illumination. By means of 
the movable tube D the amount of illumination and the angle of 



310 



I. L. NIXON 



[J. S. M. P. E. 



incidence may be regulated. A very smooth surface will be best 
illuminated by light from a small angle while a rough surface requires 




FIG. 8. Metallurgical microscope. 

higher angular illumination. The length of fibers, how they are ar- 
ranged, and how the filler and the coating has been applied may all 
be easily studied under this kind of illumination. 




FIG. 9. Complete metallographic equipment. 



Furthermore, since the tube is graduated it is possible to record the 
exact setting and to return time and again to the same illumination. 



Mar., 1932] 



OPTICAL INSTRUMENTS 



311 



Fig. 5 shows this microscope as it is now commercially made, with 
an observation tube which may be withdrawn so the light passes on 
through the other tube to the camera for making photomicrographs. 
Fig. 6 shows such a photomicrograph. Provision is made for the 
making of stereophotographs if desired. 

A number of other modifications of the standard microscope or 




FIG. 10. Path of light for large 
metallographic equipment. 



special accessories are made use of in the film laboratories in more or 
less highly specialized investigations, among which is a device known 
as a dark ground illuminator for the illumination of colloidal particles, 
a microscope with accessories for producing polarized light by means 
of which strains may be detected in crystals and film base, and a 




FIG .11. Photomicrograph of steel . 




FIG. 12. Photomicrograph of steel. 



micromanipulator, by means of which individual crystals or particles 
of impurities may be isolated and submitted to all kinds of treatment. 
While many materials may be satisfactorily illuminated with 
transmitted light there are many that are opaque and, consequently, 
must be illuminated from the top. In the case of low-power equip- 
ment this may be by means of light directed downward, striking the 



312 



I. L. NIXON 



[J. S. M. P. E. 



object at an angle, in other words, flood lighted ; while in the case of 
high-power equipment the objective works so close to the object 
that it is no longer possible to illuminate in such a manner and then 
one must resort to what is generally known as a vertical illuminator. 

Fig. 7 is a diagram showing the path of light of such a device. This 
vertical illuminator is inserted between the end of the body tube and 
the objective. In one side of the mounting is an opening usually 
fitted with a small condensing lens. A small concentrated beam of 
light from a suitable light source enters through this aperture and is 





FIG. 13. Wide field binocular 
microscope. 



FIG. 14. Comparison microscope. 



reflected 90 degrees downward, either by a clear glass reflector or by a 
prism, through the objective lens onto the specimen; and since the 
rays of light are striking normally to the surface of the specimen they 
will be reflected directly back along their original path. This is as- 
suming that the specimen is fairly well polished. If a clear glass 
reflector is used, a portion of the returning light passes through the 
glass to the eyepiece. If using a prism as the reflecting medium, it 
must be mounted off the center of the optical axis; the light then 
passes down through one side of the objective and back through the 
other, past the prism and on through to the eyepiece. 



Mar., 1932] OPTICAL INSTRUMENTS 313 

This kind of illuminator is a part of all metallurgical microscopes 
of which there are two general types. The first one, Fig. 8, is essenti- 
ally the same as the regular microscope except that it is fitted with the 
vertical illuminator and usually is without substage condenser, but it 
has the stage movable vertically by rack and pinion. This is neces- 
sary to bring the object into focus without changing the position of 
the vertical illuminator with relation to the light source after it has 
once been aligned and centered. Such a microscope as this has wide 
use as a routine instrument in the industrial laboratory. The other 
type of metallurgical microscope is that shown in Fig. 9. 

This is known as an inverted microscope, the stage being at the 
top, the specimen placed with its polished side down and the illumina- 




FIG . 1 5 . Photomicrograph of two 
pieces of textile as seen through the 
comparison microscope. 

tion coming up from underneath by means of a vertical illuminator, 
as will be seen from the diagram of the path of light in Fig. 10. 

The illuminating system of this instrument is mounted on a base 
with the microscope, so that both units may be very carefully and 
permanently aligned and centered, a point which is very essential 
when working at high magnifications. This microscope is almost 
always sold in conjunction with the camera as shown, the combination 
then being known as a metallographic equipment. It is common 
practice with such equipment to make photographs at 2000 or 3000 
diameters, and they have been made as high as 15,000 diameters. 

The value of such equipment to the manufacturer and user of 
steel, brass, copper, and all metal alloys is beyond estimation. Suffice 
it to say that if it were not for the microscope we probably would not 



314 1. L. NIXON [J. S. M. p. E. 

have the high-speed motor cars, aeroplanes, etc., that we have today. 
The chemical analysis will determine whether the correct percentage 
of the different constituents has been maintained or not, But that 
does not tell whether a piece of steel will be suitable for the purpose 
for which it has been intended or not ; such can be told only by deter- 
mining the crystalline formation after the various heat treatments, 
rolling, drawing, etc. 

A piece of steel, for instance, of a given mixture will have a very 
definite crystalline structure following certain treatments which the 




FIG. 16. Special microscope for wax records. Photograph by 
courtesy of Bell Telephone Laboratories. 

trained metallurgist can at once recognize. So he takes a small 
sample, grinds, polishes, and etches one surface, checks it on the micro- 
scope and many times will photograph it for record purposes. Such 
photographs might look like Figs. 11 and 12. 

The first shows a steel heated to a certain point of the treatment and 
the other a steel carried beyond this point in the hardening process. 

Probably one of the most useful instruments for general use around 
the laboratory or the factory is a wide field binocular microscope. 
(Fig. 13.) 



Mar., 1932] 



OPTICAL INSTRUMENTS 



315 



This, as its name implies, is arranged for binocular vision with a 
large field, and the image produced is an erect one, so that in working 
with materials, dissecting, etc., all movements maybe naturally made. 
Furthermore, one sees naturally, that is, stereoscopically. Its most 
useful range is at from magnifications of approximately IX to 3QX. 
Because of the large field and third dimension it is most useful in 
examining small parts, machine surfaces, and raw materials. 




FIG. 17. Special microscope for setting 
width of light valve. Photograph by 
courtesy of Bell Telephone Laboratories. 



In a paper presented by O. E. Conklin 1 at the 1930 Fall Meeting of 
the Society, the applications of the comparison microscope in the film 
industry are set forth in detail. Such a microscope has two objectives 
which focus on two objects to be compared, and a prism which brings 
their images together so that they can be seen in a single eyepiece side 
by side. (Fig. 14.) 

In a film laboratory, either manufacturing or processing, a compari- 
son microscope is invaluable because one is always wanting to com- 



316 



I. L. NIXON 



[J. S. M. P. E. 



pare two things, one of which may be a standard. Two films may be 
compared for general graininess, or tint, papers for surface finish, etc. 
(Fig. 15.) 

For the details of its application I refer you to Mr. Conklin's paper 
in which he also describes how this instrument had been used as the 
basic unit in the construction of a picture comparator, a sound track 
photometer, a graininess comparator, and a perforation comparator. 

This paper of Mr. Conklin's shows most conclusively how with a 




FIG. 18. Toolmaker's microscope. 

little ingenuity a single basic microscope may be made to serve in a 
number of important control steps leading to the uniformity and 
general efficiency of the film. 

In making sound records on wax a microscope becomes indispen- 
sable to check the performance of the cutting needle and for that 
purpose there has been employed a body tube with objective and eye- 
piece mounted on an arm to be swung over the record. A special 
vertical illuminator has been employed for proper illumination and 
an inclined eyepiece for greater convenience in viewing. (Fig. 16.) 



Mar., 1932] 



OPTICAL INSTRUMENTS 



317 



The setting of the width of the slit on the light valve of the Western 
Electric sound recording unit necessitates the use of a microscope 
to which a special holder is attached for holding the valve unit. 
(Fig. 17.) This equipment is provided with an optical system having 
a magnification of 100 and a scale in the eyepiece with ten spaces, 
each space representing 0.001" on the specimen. 

In modern shop practice it becomes essential to work to much closer 
limits in making screw threads, gears, cams, cutting and forming tools, 
and the mechanic is often confronted with checking curved forms 
whose contour is almost impossible to control without recourse to one 




FIG. 19. Contour measuring projector. 

or the other of two optical devices, one known as a toolmaker's 
microscope and the other as a contour projector. (Fig. 18.) 

The toolmaker's microscope is provided with an illuminating system 
producing a parallel beam of light passing the specimen and an 
objective specially corrected for use with parallel pencils of light. 
The stage has a micrometer movement of one inch in two directions at 
90 degrees to each other with graduated drums reading at a tenth of a 
thousandth of an inch. 

The eyepiece may be with a simple cross-hair against which settings 
of the image may be made, or it may be what is known as a protractor 
eyepiece with cross-hairs adjustable for angles of from 40 to 70 
degrees. 



318 



I. L. NIXON 



[J. S. M. P. E. 



With such an instrument the angle, lead, and pitch diameter of 
screw threads may be easily and accurately measured, and in addition 
it has been used particularly in the motion picture industry for check- 
ing the spacing and size of perforations, checking the tools with which 
the perforations are made and the slit on the Western Electric light 
valve. 

The contour projector, as its name implies, is a projector of contour 
forms, screw threads, gears, cutting tools, etc. (Fig. 19.) 

There are a light source, a special object holder, and a projection lens 
producing the image either on a distant screen or upon the chart 
attached to the stand depending on the size of object being in- 





FIG. 20. Path of light for contour measuring projector. 

spected. (Fig. 20.) Here again we have an optical system resem- 
bling the typical motion picture projector, except that the beam il- 
luminating the object is made up of parallel pencils. 

A five-ampere arc lamp is usually used as the light source with an 
aspheric condenser in an adjustable mount which may be set to il- 
luminate approximately a 2-inch area for large objects, and a supple- 
mentary condenser with diaphragm for use with small objects and 
high magnification objectives. 

The object may be held either between centers, in V blocks or, in 
the case of gears, on studs and special holders for special forms. The 
object holder has a forward and backward movement for focusing, and 
a vertical and transverse movement* for moving the object into the 



Mar., 1932] 



OPTICAL INSTRUMENTS 



319 



field of view. These two movements may be fitted with micrometer 
screw and drum for measuring distance, lead of screw threads, etc. 

In checking screw threads it is customary to use a special chart 
against which the thread outline is readily checked for angle, size, and 
general correctness of form. (Fig. 21.) 

A plate holder or special paper holder may be substituted for the 
chart and photographs made for record purposes. 

The contour of gear tooth form may be checked against a master 




FIG. 21. Photograph of commercial thread and thread chart. 



drawing for size and exact form or two gears may be mounted en- 
meshed and slowly rotated, and their exact rolling action carefully 
studied. The silent transmission gear systems on the present-day 
automobile are a result of the study of gear action with such a device. 
Many tool departments are using such a device for checking the 
contour of cutting tools against a master template before turning them 
over to the operating departments. 

Several manufacturers in the motion picture industry are using such 
equipment for controlling the accuracy of their mechanical parts, and 



320 



I. L. NIXON 



[J. S. M. P. E. 



no doubt the silent mechanisms of today are largely the result of a 
study of the parts involved on such a projector. Spacing and shape 
of film perforations are also being controlled by such a device. In 
measuring the spacing, one edge of a perforation is carefully lined up 
with a target on the chart, then the carrier is moved over until the edge 
of the next perforation is in line with the target. Then, with the 
micrometer screw and drum, the amount of movement or the spacing 
can be easily checked to a tenth of a thousandth of an inch. The 
radii of the corners can be easily checked against a master outline so 




FIG. 22. Densitometer. 

that the accuracy of the die and the amount of wear may be quickly 
determined. 

Up to this point we have been considering the type of equipment by 
means of which material things might be examined, measured, etc., 
but there is another group of optical instruments that are quite as 
important to the industry and their use marks in many instances the 
high state of achievement. Such instruments are photometers, and 
spectrometers in a broad sense. 

A photometer is an instrument which has for its purpose the 
measurement of light intensity. There are many kinds ranging 
from the portable kind, known as a luminometer for approximate 



Mar., 1932] OPTICAL INSTRUMENTS 321 

measuring of lumens or foot candles for screen illumination, to the 
highly specialized type for close measurement of film density. 

In designing optical systems for projector or lighting units the 
amount of illumination and its distribution are checked by a photome- 
ter, and the highly efficient systems are largely due to the ability 
to record accurately their performance with some type of photometer. 
The same thing applies to the study and development of light sources. 

We are indebted to Martens, a German physicist, for the conception 
and development of the polarization type of photometer, which has 
been embodied in a number of special instruments, an example of 

eye Poinf \ 

Scale Lens Scale Lens 

> / 

Prism ~ 
noiyzet 

Scale 



/ 
Bi- Prism 



Apertures 

opal 
GIQS.S 
DISC 



FIG. 23. Path of light of densitometer. 

which is one known as a densitometer. Its purpose is, as its name 
implies, to measure the density of the exposed portions of film. 

Fig. 22 illustrates a densitometer developed along lines suggested 
by the Bell Telephone Laboratories, with which extensive study of 
sound track density has been made. While this incorporates the 
general principle of the Martens photometer there are a number of 
modifications and the general design of the instrument has been 
around the requirements necessary for easy and accurate study of 
sound track densities. 

Fig. 23 shows the path of light of this densitometer. In this de- 
sign the two entrance pupils have been removed from the body of the 



322 I. L. NIXON [J. S. M. P. E. 

instrument and placed on the stage over which the film travels and 
are diffusely illuminated. 

The upper portion of the photometer consists of a Wollaston prism, 
a bi-prism, and an analyzing prism, with suitable lenses to produce a 
photometric field, one-half of which is illuminated by light from the 
aperture over which the film travels and the other by light from the 
clear aperture. The rotatable analyzing prism is provided with a six- 
inch divided circle, on which are engraved both density and trans- 
mission scales. These scales occupy two oppositely located 45-de- 
gree sectors, and are designed so that either transmission or density 
may be read from the same setting, the former on the left and the 
latter on the right. The scales are read by means of two magnifiers 
mounted on the eyepiece tube of the photometer. The accuracy of 




FIG. 24. Spectrophotometer. 

reading is within 3 per cent at any point in the scale and the maximum 
density which can be read is approximately 3.5. 

Other still more highly specialized instruments, known as micro- 
densitometers, have been made in very limited number by European 
manufacturers for examination of very small areas; usually used in 
the study of spectral lines, but have been used in the sound research 
laboratories for the study of individual vibrations on the sound track. 

Another specialized piece of photometric apparatus which has been 
of decidedly important usefulness to the industrialist has been the 
Spectrophotometer. The Spectrophotometer is an instrument by 
virtue of which any controllable beam of light may be split up into 
its spectral components; that is, into the component colors of light 
tending to compose it, and to measure the comparative proportion of 
each wavelength or color present in this light. It is possible by this 



Mar., 1932] OPTICAL INSTRUMENTS 323 

instrument to identify or to make a record of the color characteristics 
of a substance whether it be a reflector of colored light or, let us say, a 
piece of colored fabric, or whether it transmits colored light, for 
instance, a piece of colored glass. (Fig. 24.) 

The spectrophotometer consists of three essential parts: first, the 
source of light, in which must reside all possible colors or wavelengths 
within the visible spectrum. This condition is admirably met by the 
incandescent lamp. The second element is a spectroscope-like piece 
of apparatus, by virtue of which light entering it may be broken up 
into its component colors; and to do this quantitatively, that is, so 
that one may be able accurately to identify which color the instru- 
ment is transmitting at a particular time. The third element is a 
photometer, the function of which is to measure the intensity of the 
one-colored light being transmitted at the instant the measurement is 
taken. 

The spectrophotometer is fundamentally a kind of comparator as 
differentiated from an instrument which makes absolute measure- 
ments, but since the intensity of light originating in the incandescent 
lamp can be held practically constant by controlling its impressed 
electric current, and because of certain characteristics inherent in the 
design of the apparatus, the basis of comparison is always unity, so 
that the reading of the instrument may be reduced directly to per- 
centage of transmission of light of any given particular wavelength. 

Such an instrument is widely used by dye makers, paint manu- 
facturers, and all people who have anything to do with the specifica- 
tion of color. The dyes used in tinting film must be or should be 
checked spectrophotometrically during development to ascertain 
positively the desired color, and during manufacture to guarantee 
uniformity of product. Of course, the filters used in all processes of 
colored photography must be carefully and painstakingly studied with 
such apparatus if satisfactory results are to be obtained in their use. 

It is safe to say, therefore, that the achievement of color photog- 
raphy up to the present time and its further perfection will be due to 
the availability of spectrophotometric apparatus. 

REFERENCE 

^ONKLIN, O. E.: "Some Applications of the Comparison Microscope in the 
Film Industry," /. Soc. Mot. Pict. Eng., XVI (Feb., 1931), No. 2, p. 159. 



PHOTOGRAPHIC SENSITOMETRY, PART IV* 
LOYD A. JONES** 



The following is the fourth and final installment of Mr. Jones' paper on sensi- 
tometry, which, due to its length, was presented in part on three consecutive days at 
the Spring, 1932, Meeting of the Society at Hollywood, Calif. The preceding 
installments appeared in the JOURNALS of October and November, 1931, and January, 
1932. The paper deals in a tutorial manner with the general subject of sensitometry , 
its theory and practice. 

OUTLINE 

I. Introduction. 

04) Definition. 

(5) Scope of field. 

(C) Applications. 

(Z>) The characteristic ZMog curve. 

II. Sensitometers. 

(A) Light sources. 

(1) Historical resume. 

(a) Natural light (sunlight, skylight, etc.). 

(6) Activated phosphorescent plate. 

(c) British standard candle. 

(d) The Hefner lamp. 

(e) The Harcourt pentane standard. 
(/) The acetylene flame. 

(g) Electric incandescent lamps. 

(2) Spectral composition of radiation. 

(a) The spectral emission curve. 
(6) The complete radiator. 

(c) Color temperature of sources. 

(d) Effect of color temperature on sensitivity values. 

(3) Modern standards of intensity and quality. 

(a) Acetylene flame plus dyed gelatin filter. 
(6) Acetylene flame plus colored glass filter. 

(c) Acetylene flame plus colored liquid filter. 

(d) Electric incandescent plus colored filters. 

(4) The international unit of photographic intensity. 

(B) Exposure modulators. 

(1) Intensity scale instruments. 

* Presented at the Spring, 1931, Meeting at Hollywood, Calif. 
** Kodak Research Laboratories, Eastman Kodak Co., Rochester, N. Y. 
324 



PHOTOGRAPHIC SENSITOMETRY 325 

(a) Step tablets (7 variable by finite increments). 
(6) Wedge tablets (/ variable by infinitesimal incre- 
ments). 

(c) Luther's crossed wedge tablet. 

(d) Tube sensitometer. 

(e) Optical systems with step diaphragms. 

(/) Optical systems with continuously variable dia- 
phragms. 
(2) Time scale instruments. 

(a) Exposure intermittent. 

Finite exposure steps (discontinuous gradations). 
Infinitesimal exposure steps (continuous grada- 
tions). 

(b) Exposure non-intermittent. 

Finite exposure steps (discontinuous gradations). 
Infinitesimal exposure steps (continuous grada- 
tions). 
HI. Development. 

04) Developers. 

(1) Standards for sensitometry. 

(a) Ferrous oxalate. 

(b) Pyro-soda. 

(c) -Aminophenol. 

(2) Standards for control of processing operations. 
(B} Temperature control. 

(C) Development technic. 

(1) For standardized sensitometry. 

(2) For control of processing operations. 

IV. The measurement of density. 

04) Optical characteristics of the image. 

(1) Partial scattering of transmitted light. 

(2) Diffuse density. 

(3) Specular density. 

(4) Intermediate density. 

(5) Relation between diffuse and specular values. 

(6) Effective density for contact printing. 

(7) Effective density for projection. 

(8) Color index. 

(B) Fog and fog correction. 

(1) Source of fog. 

(a) Inherent fog. 
(6) Processing fog. 

(2) Fog correction formulas. 

(C) Densitometers. 

(1) Bench photometer, 
(a) Rumford. 
(6) Bunsen. 



326 LOYD A. JONES IJ. S. M. p. E. 

(c) Lumer Brodhun. 

(2) Martens polarization photometer. 

(a) Simple illuminator. 

(6) Split beam illuminator. 

(3) Integrating sphere. 

(a) For diffuse density. 

(6) For diffuse and specular density. 

(4) Completely diffused illumination. 

(a) For diffuse density. 

(5) Specialized forms. 

(a) Furgeson, Renwick, and Benson. 
(V) Capstaff-Green. 

(c) High-intensity (Jones). 

(d) Density comparators. 

(6) Physical densitometers. 

(a) Thermoelectric. 
(6) Photoelectric. 
(c) Photovoltaic. 

V. Interpretation of Results. 

04) Speed or sensitivity. 

(1) Threshold speed. 

(a) Scheiner speed numbers. 

(b) Eder-Hecht. 

(2) Inertia speeds. 

(a) H & D scale. 
(&) Watkins scale. 

(c) Wynne scale. 

(3) Luther's crossed wedge method. 

(4) Minimum useful gradient. 

(B) Gamma infinity, y^. 

(C) Velocity constant of development, K. 

(D) Time of development for specified gamma. 

(1) T d (y = 1.0). 

(E) Latitude, L. 
(/O Fog, F. 

VI. Spectral Sensitivity. 

(A) Dispersed radiation methods. 

(1) Monochromatic sensitometers 

(2) Spectrographs. 

(a) Ordinary. 
(6) Glass wedge. 
(c) Optical wedge. 

(B) Selective absorption methods. 

(1) Tricolor. 

(2) Monochromatic filters. 

(3) Progressive cut filters. 



Mar., 1932] PHOTOGRAPHIC SENSITOMETRY 327 

VI. SPECTRAL SENSITIVITY 

Any treatment of the subject of sensitometry, especially in these 
days of great popularity of panchromatic materials, cannot be con- 
sidered as complete without a discussion of the various methods used 
for the measurement and specification of the sensitivity of photo- 
graphic materials to radiation of different wavelengths. This 
characteristic is usually referred to as spectral sensitivity. A knowl- 
edge of the way in which sensitivity is distributed throughout the 
spectrum, including the ultra-violet and the "infra-red as well as the 
visual regions, is of great importance from both the practical and 
the theoretical points of view. The rendition of color by a photo- 
graphic process is determined largely by the spectral sensitivity of the 
negative material. For instance, it is well known that the ordinary 
blue-sensitive materials, red, orange, and yellow, are rendered in about 
the same tone value as black, while in the case of some panchromatic 
materials, which have been rendered very sensitive to the longer wave- 
lengths of visible radiation, these same colors may be rendered "as 
almost white. The correct rendering of colored objects on the black 
to white tone scale, which represents the entire discrimination gamut 
of the photographic process, is conditioned almost entirely by the 
spectral sensitivity of the material. It is evident, therefore, that a 
knowledge of this characteristic of photographic materials is of great 
importance wherever orthochromatic rendering of colored objects is to 
be considered. 

Many workers in the field of photography have studied this problem 
of spectral sensitivity and, in fact, its investigation dates back almost 
as far as the beginning of sensitometry. As early as 1882 Abney 106 
studied the spectral sensitivity of photographic materials by exposing 
them to dispersed radiation in a spectroscope and by plotting densi- 
ties, as determined by visual estimation, against wavelength ob- 
tained a curve of spectral sensitivity. Abney later improved the 
method until finally, in 1888, 107 he suggested a technic which involved 
impressing on the same plate an exposure to a spectrum and a series of 
known exposures to white light. The opacities of the spectrum ex- 
posures were then measured and interpolated between those of the 
white light exposures. He thus obtained a curve showing the equiva- 
lent spectral intensities for various wavelengths. 

The exposures of the photographic material to dispersed radiation 
afforded by an instrument of the spectroscopic type gives information 



328 LOYD A. JONES [J. S. M. p. E. 

. which, while it is undeniably complete, nevertheless is not conveni- 
ently expressible by simple numerical values but must be shown in 
graphic form. Many methods have been developed, therefore, which 
employ not spectroscopically dispersed radiation but spectral bands, 
more or less narrow, isolated by selective absorption. This may be 
accomplished either by using transmitting materials such as colored 
glass, dyed gelatin, etc., or reflecting materials such as pigment coated 
surfaces. One of the earliest of such color sensitometers was also 
devised by Abney. 108 Since that time almost numberless methods 
have been proposed for the measurement and expression of color 
sensitivity. No attempt will be made at this time to give a complete 
bibliography of the subject but a few references to some of the more 
recent work may be of interest. Leimbach 109 has made a systematic 
study of the spectral energy distribution for five different emulsions 
in the region between 450 and 700 imx. He found the maximum 
sensitivity to occur in the spectral region corresponding to 450 mju. 
Luckiesh, Holladay, and Taylor 110 have published sensitivity curves of 
four emulsions indicating a maximum sensitivity near 450 mju. 
Otashiro 111 found maximum sensitivity at about 465 m/z, the sensi- 
tivity decreasing uniformly through the blue and violet. Helmick 112 
(using an emulsion of the ordinary blue- sensitive type) has measured 
the average number of quanta required to make a silver bromide grain 
developable by radiation at various isolated wavelengths ranging 
from 253.7 to 549.0 m/x. He found that the least number of quanta 
per grain are required at wavelength 549 and the maximum number at 
wavelength 253.7 imx. More recently Harrison 113 has published 
results showing the relation between sensitivity and wavelength for 
six different photographic plates in the region between 200 and 
450 mju. His results indicate that sensitivity is practically constant 
for wavelengths greater than 250 imx, decreasing rapidly for wave- 
lengths less than 250 m^i. He also shows the relation between 
contrast (gamma) and wavelength. 

All methods used for obtaining information as to the spectral sensi- 
tivity of photographic materials involve the isolation of more or less 
narrow spectral bands and then observing either qualitatively or 
quantitatively the response produced when the material is exposed 
to these more or less homogeneous radiations. For this purpose, a 
wide variety of spectral instruments monochromatic sensitometers, 
spectrographs, tricolor tablets, ratiometers, color charts, and filter 
assemblies have been devised. 



Mar., 1932] 



PHOTOGRAPHIC SENSITOMETRY 
DISPERSED RADIATION METHODS 



329 



The more refined and elegant methods involve the dispersion of 
radiation by some suitable element, such as a prism or diffraction 
grating. In this way radiation of high spectral purity may be ob- 
tained to which the photographic material may be exposed. Instru- 
ments of this type may be divided, for the sake of convenience, into 




FIG. 51. 



Diagram of optical system of monochromatic 
sensitometer. 



two general classes: the first including those instruments in which 
the photographic material is exposed to the entire spectrum at the 
same time; the second including those where a single very narrow 
band of practically homogeneous radiation is isolated, to which the 
photographic material is exposed in a manner similar to that used in 
the sensitometers already described in the earlier sections of this 
paper. While it is impossible to draw a strict line of demarcation, 



330 



LOYD A. JONES 



[J. S. M. P. E. 



the term "monochromatic sensitometer" is usually applied to instru- 
ments of the second class, and the term "spectrograph" is used in 
reference to those of the first class. 

Monochromatic Sensitometer s. The optical system employed for 
isolating monochromatic radiation of high spectral purity as used in a 
monochromatic sensitometer described by Jones and Sandvik 114 is 
shown in Fig. 51. This consists essentially of two quartz mono- 
chromatic illuminators, A and B. The radiation emerging from the 
exit slit of the first monochromatic illuminator, A, passes into the 




FIG. 52. The sector disk of monochromatic sensitometer. 

entrance slit of the second monochromatic illuminator, B. In this 
way practically all the stray radiation may be eliminated so that the 
radiation emerging from the exit slit of the second monochromatic 
illuminator consists entirely of the wavelength as indicated by the 
wavelength drums of the two instruments. Great emphasis must be 
laid on the necessity of obtaining high purity for work of this kind, 
especially if it is desired to work in those spectral regions where the 
amount of energy obtainable from the light source used is relatively 
low as compared with that present in other spectral regions. For a 



Mar., 1932] 



PHOTOGRAPHIC SENSITOMETRY 



331 



complete discussion of this subject and of the relatively great errors 
which may be introduced by failure to eliminate all stray radiation, 
the reader is referred to the more complete discussion of the subject 
by Jones and Sandvik (loc. cit.). 

After having been isolated, the homogeneous radiation is allowed to 
fall on the photographic plate and, by means of a suitable mechanism, 
the time of exposure is varied in a known manner. It is usually 
difficult to obtain a quantity of monochromatic radiation so that it 
may be spread over a sufficient area of the photographic material to 
permit the use of the conventional type of exposure time controlling 
elements. It is usually necessary, therefore, to illuminate a relatively 





FIG. 53. Schematic diagram of monochromatic sensitometer. 

small area of the photographic plate and to make the exposures of the 
various steps on the sensitometric strip one after the other, varying 
the time factor of exposure in the desired manner. The method 
adopted by Jones and Sandvik (loc. cit.) employs a sector disk of 
special type in which the apertures are arranged spirally around the 
axis of rotation, the entire disk being moved laterally, while rotating 
at a uniform angular velocity. The structure of this disk is shown 
in Fig. 52 and the arrangement of the essential parts of the exposing 
mechanism is shown in Fig. 53. The shutter disk as shown is keyed 
to the shaft carried by a movable bearing sliding between the ways 
MM. The rotation of the lead screw H (driven by the same shaft 
which imparts rotational motion to the shutter disk) moves the 



332 



LOYD A. JONES 



[J. S. M. P. E. 



shutter disk laterally while it is being rotated. Mounted on the shaft 
carrying the shutter disk is a cam disk which carries a series of thirteen 
cam elements. As these cam elements rotate with the cam disk they 
close the electrical contact, 7, at definitely predetermined intervals. 
The closing of this contact energizes the solenoid, Q, which, through 
a suitable escapement, moves the photographic plate forward one 
step during the time when the opaque element of the shutter disk 
occupies a position in the path of the exposing radiation. By utilizing 
the spiral arrangement of the apertures the maximum exposure time 



^8 



ZA 



0.8 



0.4 





"2.0 1A 



I.& 



O.Z. 
I_OG> 



O.fe 



1/2. 



\.Q> 



FIG. 54. 



0.0 04 O 
/cm*) 

D-log E curves obtained by exposing to monochromatic radiation of 
wavelength 350 m/z. 



corresponds to an angular rotation of 720 degrees of the shutter disk. 
In this way twelve exposures increasing by consecutive powers of two 
are obtained, thus giving a range of exposure times from 1 to 2048. 
In this way an exposed sensitometric strip of the conventional type is 
obtained. 

By placing a thermopile in the exposure plane of the sensitometer, 
that is, in the same position as that occupied by the photographic 
material during exposure, the energy flux density of the monochro- 
matic radiation may be measured. Since this energy value is usually 
relatively small, it is necessary to use a thermopile-galvanometer 



Mar., 1932] 



PHOTOGRAPHIC SENSITOMETRY 



333 



combination of high sensitivity, and great care must be exercised in 
making the energy measurements in order to obtain reliable and 
precise results. The determination of these energy values with the 
required precision is perhaps the most difficult step in the process of 
obtaining absolute values of spectral sensitivity. 

The monochromatic sensitometer thus gives an exposed strip of the 
conventional type except that the exposure values are expressed in 
terms of energy (ergs) per unit area. This is developed under fixed 
conditions, the densities read, and a curve plotted which, of course, is 




O.O O.4 

LOG E (ERGS. /cm 1 ) 



1.2. 



FIG. 55. D-log E curves obtained by exposing to monochromatic radiation of 

wavelength 450 m/z. 

the density-log E characteristic for that particular wavelength of 
radiation. In making a complete study of the spectral sensitivity 
of the material it is necessary to expose several sensitometric strips at 
each wavelength. These strips are then subjected to a series of 
increasing development times, and in this way a complete family of 
characteristic curves is obtained for each wavelength. These may 
then be interpreted in the usual manner yielding the usual time of 
development-gamma curve, time of development-fog curve, etc., for 
each wavelength. By proceeding in this manner throughout the 
spectrum, exposing a set of strips at a sufficient number of different 



334 



LOYD A. JONES 



[J. S. M. P. E. 



wavelengths, a complete set of data is obtained from which the various 
spectral response curves may be plotted. 

In Figs. 54, 55, 56, and 57 are shown typical families of D-log E 
characteristic curves obtained by exposing a panchromatic material at 
wavelengths 350, 450, 600, and 700 m//, respectively. Careful 
comparison of these curves will show that the wavelength of radiation 
used in making the exposures has a profound influence upon the 
general shape of these curves. The development times used were 2.5, 
3.5, 5, and 7.5 minutes, and it is quite evident from an inspection of 



2.4 



2.0 



0.8 



0.4 





Z.O 1.4 



O.TL O-Q> O.O 0.4 O.fc 1.7. i. 
LOG I 



FIG. 56. D-log E curves obtained by exposing to monochromatic radiation of 

wavelength 600 m/*. 

these figures that gamma rises to a higher value in case of the exposures 
made to the longer wavelengths than in case of those made to the 
shorter wavelengths. While photographic materials differ to a certain 
extent among themselves in their response to radiation of different 
wavelengths, the effect mentioned is rather typical, although, of 
course, there may be some exceptions. 

In Fig. 58 are plotted the complete gamma-wavelength curves for 
the four times of development as mentioned previously. It is 
interesting to note that gamma increases from a minimum value at 
the short wavelength end to a maximum at approximately 550 m/i, 



Mar., 1932] 



PHOTOGRAPHIC SENSITOMETRY 



335 



decreasing from this point to a minimum at about 650 m/*, after 
which it rises again as wavelength increases. There is relatively little 
variation in gamma for the shortest time of development but if time 
of development is increased the dependence of gamma upon wave- 
length of the exposing radiation becomes much more marked. 

In Fig. 59 is plotted a group of curves, one for each of the different 
wavelengths as indicated, all of these being obtained by a single time 
of development, namely, 5 minutes. It will be noted that there are 
characteristic differences depending upon the wavelength of the 



Z.4 



t\& 

ifl 



0.8 



0.4- 




FIG. 57. ZMog E curves obtained by exposing to monochromatic radiation 
of wavelength 700 m/z. 



exposing radiation. For instance, the curves resulting from ex- 
posures to the shorter wavelengths show lower maximum densities 
and appreciably greater latitude than those obtained by exposures to 
the longer wavelengths. In plotting these curves their relative 
positions with respect to the log exposure axis must not be taken as 
indicative of the sensitivity of the material to the various wavelengths 
of radiation. They are assembled from left to right with increasing 
values of wavelength in a convenient manner to show the relative 
shapes. Their actual relationships to the log exposure scale are given 
by the values of log exposure as indicated at the intersection of the 



336 



LOYD A. JONES 



[J. S. M. P. E. 



straight line portion of the characteristic curve with the log exposure 
axis. These values are the log exposures corresponding to the respec- 
tive inertia points. 

The problem of expressing sensitivity must now be considered, and 
it is obvious that the shape of the wavelength sensitivity curve will 
depend profoundly upon the manner in which sensitivity is defined. 
In expressing spectral sensitivity it is necessary to depart from the 
method which has already been defined for expressing the speed or 
sensitivity of a photographic material. It will be recalled that speed 



zo 



r 

08 



06 
0.4 
OZ 




feftO 



FIG. 58. Gamma-wavelength curves for various times of development as 

indicated. 

value to heterogeneous radiation (white light) is defined, for ordinary 
sensitometric purposes, in terms of exposure where exposure is 
expressed in meter candle seconds. Now, the meter candle is a unit of 
illumination and is measured visually or, even if measured by some 
radiometric or physical method of photometry, is referred to as the 
unit of luminous intensity, the international candle. For the expres- 
sion of monochromatic sensitometric results it is obvious that this 
method is quite useless. For instance, let it be assumed that mono- 
chromatic radiation of wavelength 350 m/i is used. The eye is 
entirely insensitive to radiation of this wavelength and hence the 



Mar., 1932] 



PHOTOGRAPHIC SENSITOMETRY 



337 



luminous intensity, that is, candle power, of such radiation will be zero 
no matter how great the radiant flux or radiant intensity (expressible 
in units of radiant flux or radiant flux density) may be. For the 
purpose of monochromatic sensitometry, it is necessary, therefore, to 
express exposure in terms of energy units, and the unit most usually 
used is the erg. Since the photographic material integrates more or 
less perfectly the energy which falls upon it over a period of time, it is 
necessary of course to include the time factor, and in expressing 
photographic exposure in energy units it is necessary to multiply the 
rate at which energy falls upon the surface (radiant flux density) by the 
time during which the exposure persists. Exposure, therefore, must 
be expressed in terms of ergs (or other suitable energy units) per unit 



20 



t 1.6 



* 



ZL 



I4O 1^4 I 40 1.40 ZOO 

LOG EXPOSURE. 



FIG. 59. .D-log E curves for the various wavelengths as indicated; develop- 
ment time 5 minutes. 



area. The abscissa values in Figs. 54 to 57, inclusive, are therefore 
in terms of log exposure where exposure is expressed in terms of ergs 
per sq. cm. We may not express the sensitivity of the photographic 
material for any particular wavelength in a manner analogous to that 
used in white light (heterogeneous radiation) sensitometry. Thus we 
may use the value of inertia, which now must be expressed in terms of 
ergs per sq. cm. as a means of deriving a sensitivity number which of 
course must be proportional to the reciprocal of the inertia value. Or 
if it is desired to use any other method of speed expression, such, for 
instance, as the exposure required to give a just perceptible density 
(threshold speed) or the exposure required to give some minimum 
gradient (gradient speed), this can be done; but it must be kept in 



338 LOYD A. JONES [J. S. M. p. E. 

mind at all times that exposure is not expressed in terms of meter 
candle seconds. 

Having available now information as to the sensitivity of the photo- 
graphic material to radiations of different wavelengths, it remains to 
consider a suitable method of expressing the spectral sensitivity. It 
is quite possible of course to compute the sensitivity at various wave- 
lengths in terms of reciprocal inertia. By plotting these reciprocal 
inertia values as a function of wavelength a spectral sensitivity curve 
will be obtained, and for many purposes such a method of expressing 
spectral sensitivity seems to be quite satisfactory. It should be 
pointed out, however, that spectral sensitivity may be expressed in 
other ways, and it is possible that some of these may be somewhat 
more useful from the practical point of view. 

Referring now to Fig. 59 it is apparent that if spectral sensitivity 
be defined in terms of the energy required to give a density of unity 
for a fixed time of development, the shape of the spectral sensitivity 
curve will be quite different from that based upon inertia. Moreover, 
if a higher density value were chosen a still further modification in 
the shape of the spectral sensitivity curve will be obtained. There 
does not seem to be any means of deciding just what mode of express- 
ing spectral sensitivity will be found most desirable from all points 
of view and, in fact, it seems very probable that the method chosen 
must depend upon the particular problem in hand. For theoretical 
purposes there is considerable argument for defining spectral sensi- 
tivity in terms of the energy required to give a density of unity when 
development for all wavelengths is carried to a gamma of unity. For 
practical purposes, however, it seems that the evaluation of spectral 
sensitivity in terms of a fixed development time is more suitable and, 
in order to discount somewhat the misleading effects of gamma varia- 
tion, it seems probable that the determination of the energy per unit 
area required to give a density of unity for a fixed time of development 
is most satisfactory as a mode of expressing spectral sensitivity. The 
most suitable development time is probably that which produces on a 
sensitometric strip exposed to white light a gamma approximately 
equal to that at which the material is usually developed in practice. 
In Fig. 60 is shown a spectral sensitivity curve determined in this 
manner. This is for high speed panchromatic motion picture film, 
the development time used being that which gives a gamma of 0.7 on 
a white light sensitometric strip. 

It should be borne in mind that the spectral sensitivity curve, when 



Mar., 1932] 



PHOTOGRAPHIC SENSITOMETRY 



339 



plotted in accordance with the specifications given in the last para- 
graph, represents the characteristics of the photographic material 
itself, quite apart from any consideration of the energy distribution in 
the light source used. The curve as shown in Fig. 60 shows the 
response of this material when used with a light source emitting the 
same amount of energy at all wavelengths. In practice, light sources 
used for photography depart appreciably from this condition of an 
"equal energy" spectrum. Where it is desired to determine the 
effective response of a photographic material when used with a light 



I4O 



izo 



>ioo 



BO 



j- 



TOO 



FIG. 60. Spectral sensitivity curve for a high speed panchromatic motion 

picture film. 



source which is not emitting equal energy at all wavelengths, it is of 
course necessary to compute a new relationship which may be termed 
the photicity of the particular photographic material-light source com- 
bination. In considering the rendition of colored objects in practice 
it is very important to consider this effective response curve. 

In order to illustrate the profound influence which the spectral 
composition of the radiation may exert on color rendition, the curves 
in Fig. 61 are shown. The dotted curve, A, represents the distribu- 
tion of energy in the radiation emitted by an incandescent tungsten 
filament operating at a color temperature of 3000 degrees, which is an 



340 



LOYD A. JONES 



[J. S. M. P. E. 



efficiency frequently met with in practice. The curve B is obtained 
by multiplying, wavelength by wavelength, the ordinates of curve A 
in Fig. 61 by the ordinates of the curve shown in Fig. 60. This then 
becomes the photicity curve for a 3000-degree tungsten lamp as 
evaluated by a high speed panchromatic material. It should be 
noted that the relatively small amount of energy present in this radia- 
tion at the shorter wavelengths produces a marked decrement in the 
effective response in the longer wavelength part of the spectrum. This 
accounts for the fact that with this combination of photographic 




noo 



FIG. 61. Curve A, spectral energy curve of incandescent tungsten at color 
temperature 3000 K. Curve B, photicity curve for a 3000 tungsten lamp as 
evaluated by a high speed panchromatic material. 

material and light source, colors such as red, orange, and yellow are 
rendered on the neutral tone scale by brightnesses which are relatively 
too high as compared with their true positions on the visual brightness 
scale. 

Where a complete analysis of the spectral sensitivity characteristics 
of a photographic material is required, the foregoing methods are 
undoubtedly superior to any of the less perfect analyses which may be 
obtained by the use of spectrographic records or by the use of the 
various test chart methods relying upon selective absorption of dyes 
or pigments. These latter methods, however, are frequently much 



Mar., 1932] 



PHOTOGRAPHIC SENSITOMETRY 



341 



simpler and more rapid. In many cases they give results which are 
quite significant and for some practical purposes entirely adequate. 

Spectrographs. In instruments usually referred to as spectrographs 
the radiation from some suitable source is dispersed by means of a 
diffraction grating or prism, and the spectrum thus produced is 
allowed to fall directly upon the photographic material. The resul- 
tant spectrogram gives considerable information relative to the spectral 
sensitivity of the material. The method has the advantage of 
simplicity and rapidity. These results usually are inspected directly 
and estimates are made of the amount of sensitivity at various wave- 
lengths. It is quite possible, of course, to obtain quantitative data by 



Condenser 




FIG. 62. Optical system of wedge spectrograph. 

making densitometric measurements of the silver deposits, and under 
certain conditions this method may yield data of a high order of pre- 
cision. The usual forms of densitometers are not adapted for reading 
the densities in these spectrograms and it is generally necessary to use 
microdensitometers which are designed to measure the density of rela- 
tively small or at least very narrow elements of the spectrogram. 

It is necessary in work of this kind to know definitely the distribu- 
tion of energy incident at various points on the photographic material. 
This may be measured directly by means of a thermopile so arranged 
as to pick up a relatively narrow line element in the spectrum plane. 
The distribution of energy may in some cases be computed. This of 
course presupposes a precise knowledge of the spectral emission 



342 LOYD A. JONES [J. S. M. P. E. 

characteristics of the light source and, furthermore, a complete 
knowledge of the optical characteristics of the dispersing system, such, 
for instance, as slit width, dispersion, losses due to reflection and 
scattering within the optical system, etc. 

As stated previously, the spectrographic method is of particular 
utility where a graphic record meets all of the requirements of the 
problem. By controlling the distribution of energy incident upon 
the entrance slit, the spectrograph may be made to give directly a 
graphic representation of the effective spectral response curve of the 
photographic material and light source used. Instruments of this 
kind are usually referred to as wedge spectrographs. The distribu- 
tion of radiation on the entrance slit may be controlled by a rotating 
sector of logarithmic form placed between the light source and the 
slit of the instrument. In this way the energy incident upon the slit 
can be made to decrease from one end of the slit to the other according 
to a logarithmic law. Such rotating sectors, since they must be 
quite small, are rather difficult and expensive to manufacture, and a 
better solution of the problem is obtained by using a neutral gray 
wedge placed directly over the slit of the spectrograph as proposed by 
Mees and Wratten. 115 The construction of such an instrument is 
shown in Fig. 62. In this a diffraction grating is used which gives 
normal dispersion and it is therefore considerably more convenient 
than prism instruments which compress into a relatively small space 
the long wavelength end of the spectrum and stretch out unduly the 
short wavelength end. A suitably engraved scale plate is placed in the 
plate holder so that during exposure the sensitive surface of the photo- 
graphic material is in direct contact with this scale plate. In this 
way the wavelength scale is printed directly on the spectrogram, thus 
facilitating the reading of the results. 

In Fig. 63 are shown examples of records obtained in the wedge 
spectrograph with photographic materials having various spectral 
sensitivities. The envelopes of the light portions constitute the 
spectral response curves for the various photographic materials as 
used with a particular light source. In the case illustrated the 
quality of radiation used was approximately equivalent to that of 
noon sunlight. It should be remembered in the interpretation of 
these spectrograms that the wedge used over the slit has a linear 
density gradient and therefore the distribution of radiation along the 
slit decreases logarithmically from one end to the other. These 
envelope curves therefore are in logarithmic form, and cannot be 



Mar., 1932] 



PHOTOGRAPHIC SENSITOMETRY 



343 



compared directly with spectral sensitivity curves such as are illus- 
trated in Fig. 60 where the ordinates are relative sensitivity, not the 
logarithms of sensitivity. . 

One other fact should be kept in mind in judging these spectro- 
grams. The apparent decrease in sensitivity at the short wavelength 
end is due to selective absorption in the neutral gray glass of which the 
wedge is manufactured. So far as the author knows, all of the so- 
called neutral gray glasses depart appreciably from neutrality in the 
wavelength region below 440 imz, the absorption there being consider- 



ORDiMKRV 



ORTHOCHROMArriC 




c) 



FIG. 63. Wedge spectrograms obtained with instrument illustrated in Fig. 62. 

ably greater than throughout the rest of the visible spectrum. While 
this is inconvenient it is not particularly serious since interest is 
usually centered on the distribution of sensitivity for wavelengths 
longer than 440 mju. It is well established also that photographic 
materials differ relatively little among themselves in the distribution 
of sensitivity in the region of shorter wavelengths. The reader should 
be cautioned again to remember at all times that wedge spectrograms 
made in this manner include not only the spectral characteristic of the 
material but also the spectral emission characteristic of the source 
used in illuminating the spectrograph. 



344 LOYD A. JONES [J. S. M. P. E. 

It is possible to avoid the undesirable selective absorption charac- 
teristics of a neutral gray glass wedge by the use of a specially de- 
signed non-spherical condensing system for the illumination of the 
slit of the spectrograph. Such a condenser was proposed by Callier 116 
and a modified form of the Callier condenser has been used by 
Hansen. 117 More recently Miller 118 has published a paper in which 
an improved form of this condenser is described. Such a condenser 




FIG. 64. Wedge spectrograms obtained with Miller's optical wedge spectro- 
graph. 

involves the use of a diaphragm which may be cut so as to give any 
desired distribution of illumination on the slit of the instrument. 
For photographic purposes a logarithmic distribution is usually most 
desirable and it may be made either continuous or in the form of steps 
as required. The form adopted by Miller is that of a stepped logarith- 
mic diaphragm giving results as illustrated in Fig. 64 which represents 
the spectral sensitivity of a panchromatic, an orthochromatic, and an 
ordinary (blue-sensitive) photographic material. An inspection of 
the spectrograms in Fig. 64 will show that they carry out into the 



Mar., 1932] PHOTOGRAPHIC SENSITOMETRY 345 

short wavelength region much better than those obtained with the 
neutral gray glass wedge instrument as illustrated in Fig. 63. 

The use of a stepped wedge or diaphragm is particularly 
advantageous where it is desired to make actual density measure- 
ments and to obtain quantitative results from these wedge spectro- 
grams. The short wavelength cut-off obtainable with an instrument 
of this type is determined by the absorption of the glass lenses and by 
the distribution of energy in the source used as an illuminant. In 
case it is desired to extend the work into the ultra-violet region an 
instrument with quartz optics may be used. In order to take ad- 
vantage of this transmission, of course, it is necessary to use a light 
source emitting radiation throughout the ultra-violet. 

An inspection of wedge spectrograms yields a great deal of informa- 
tion as to the distribution of sensitivity and also some qualitative idea 
of the variation in gamma with wavelength of the exposing radiation. 
They cannot be considered as satisfactory as the determinations 
made by methods of monochromatic sensitometry, described in a 
previous section, but where it is desired to have permanent compara- 
tive records which can be obtained easily and without undue labor the 
wedge spectrogram has much to commend it. 

SELECTIVE ABSORPTION METHODS 

The spectral sensitivity of a photographic material as determined by 
the methods of monochromatic sensitometry and by the usual spectro- 
graphic technic is most conveniently and almost necessarily expressed 
graphically, the usual mode being a curve showing sensitivity as a 
function of wavelength. It is almost impossible to express the 
information relative to spectral sensitivity as derived by these methods 
in brief numerical terms. The results of course can be shown in 
tabular form in which the sensitivity values are given for certain 
selected wavelengths, but in general such a tabulation is not particu- 
larly convenient and is too complex and voluminous for practical 
purposes of classification and record. In order to obtain a more 
simple specification of color sensitivity which can be expressed by a 
few numerical values, it is frequently convenient to depart from the 
monochromatic method and determine the response of photographic 
materials to relatively broad spectral bands. This may be ac- 
complished with apparatus of the spectrographic type using, instead 
of very narrow spectral regions, broad bands, each embracing a 
relatively large proportion of the entire spectral range. If 



346 LOYD A. JONES [J. S. M. P. E. 

results of this type are desired it is usually much more convenient to 
resort to methods of selective absorption for the isolation of the de- 
sired spectral regions. Incidentally, the instrumental equipment 
required for this work is much less expensive than that for the spectro- 
graphic type. As mentioned previously, this method of obtaining a 
numerical expression for spectral sensitivity is very old, being used 
first probably by Abney in about 1895. 119 The test as devised by 
Abney consisted of a tablet made of a series of colored glasses, each 
transmitting a relatively broad spectral band and adjusted in such a 
way as to give equal illuminations. A similar method was used by 
Eder for testing the relative spectral sensitivity of orthochromatic 
plates. In his earlier work the spectrum was divided into two parts, 
one containing all wavelengths longer than approximately 495 m/z 
and the other all wavelengths shorter than this value. This wave- 
length represents approximately the long wavelength limit of the 
sensitivity of an ordinary non-color sensitized material. The values 
obtained by this method, therefore, give the ratio of the sensitivity 
due to optical sensitizing as compared to that due to the inherent 
sensitivity of the unsensitized silver halide. Later, Eder 120 divided 
the spectrum into three regions: orange-red, green, and blue- violet. 
E. J. Wall 121 also employed three selectively absorbing niters dividing 
the spectrum into three parts similar to those used by Eder. Since 
this early work almost numberless devices have been constructed and 
used, employing colored glasses, dyed gelatin, colored solutions 
or pigment coated surfaces for the isolation of more or less narrow 
spectral bands. No attempt will be made to give a complete bibli- 
ography of this subject but one or two of those methods which have 
been most extensively used will be described and discussed briefly. 

Tricolor Filters. Probably the most widely used method of this 
type involves the use of three filters having selective absorption so 
adjusted as to divide the visible spectrum into three approximately 
equal wavelength bands. As typical of such filters, the Wratten tri- 
color sets may be mentioned, and, in fact, since this set of filters has 
become almost standard throughout the world for three-color photo- 
mechanical processes, the expression of color sensitivity in terms of 
these three filters has become quite universal. In Fig. 65 are plotted 
the spectral transmission curves of the three filters in question, 
namely, Wratten No. 25 (A), red; Wratten No. 58 (B), green; Wratten 
No. 49 (C4), blue. The red filter (No. 25) transmits quite freely all 
radiation of wavelength greater than 600 m/z. It has a maximum 



Mar., 1932] 



PHOTOGRAPHIC SENSITOMETRY 



347 



transmission of 80 per cent, and hence is quite efficient as a means of 
isolating the third of the visible spectrum lying between 600 and 700 
m/z. The green filter (No. 58) has a maximum transmission at wave- 
length 520 m/z but at this wavelength it transmits only 60 per cent of 
the incident radiation. Its transmission falls rapidly on both sides of 
this point, the lower transmission limit being approximately 480 m/x 
and the upper limit 600 imx. While its total transmission computed 
on the basis of energy is relatively low, its transmission for white light 
as measured visually is relatively high since the maximum of the 
visual sensitivity lies at approximately 550 rmx. The blue filter (No. 



140 

IZO 




80 



60 



h 40 



20 



\ D 



/T\ 




300 



AOO ?>OO <bOO 

WA V E l_EI NGTTK 



TOO 



FIG. 65. Spectral transmission curves of Wratten tricolor filters. 



49) has its maximum transmission at 455 m/i at which point 
the transmission value is only 27 per cent. For both longer and 
shorter wavelengths transmission decreases rapidly, the short wave- 
length limit being approximately 360 m/x and the long wave- 
length limit 500 imz. Evaluated in terms of visual transmission 
for white light its efficiency is low, its transmission value determined 
in this manner being only 0.5 per cent. It is relatively inefficient as a 
means of isolating the third of the visible spectrum lying between 400 
and 500 m/z, but since photographic materials in general are very 
sensitive in this region its photographic transmission is quite high. 



348 



LOYD A. JONES 



[J. S. M. P. E. 



For the panchromatic materials which were in use up to one or two 
years ago, this filter had a multiplying factor of 8 which is fairly 
comparable to the multiplying factors of the green and blue filters as 
measured in terms of these older panchromatic materials. For the 
panchromatic materials recently introduced,which have a much higher 
proportion of their total sensitivity concentrated in the green and red 
regions, the multiplying factor for this filter is appreciably higher, 
being of the order of 16. 



(a) 



(b) 



O 





ul 


^ 

u 




E1/\R 


J 


u 
o 

r> 


D 
U 






1 


(D 

i 


i 


1 


(J) 


co 







^- 


iO 


pj 













FIG. 66. 



Exposure = a 8a 8a 8a 

(a) Tricolor tablet; (b) result obtained with tricolor filter method. 



In practice, sensitometric results are obtained by using a tricolor 
tablet, the structure of which is illustrated at (a) in Fig. 66. Strips 
of these filters, which are manufactured in the form of dyed gelatin, 
together with a strip ef plain (undyed) gelatin film No. 0, are 
cemented between two sheets of glass. The dimensions of this 
tablet are such that it just fits into the plate holder of a sensitometer 
of the ordinary white light type, and the strips are of such width that 
one sensitometric exposure is made through each of the four filters. 



Mar., 1932] PHOTOGRAPHIC SfiNSITOMETRY 349 

In order to obtain exposures through the tricolor filters, which balance 
fairly well with that made through the clear filter, it is customary to 
increase the time of the tricolor filter exposures so that it is eight 
times as great as that made through the clear area. 

At (&) Fig. 66 is shown a reproduction of the result obtained by 
application of this method to a panchromatic material. Since the 
relative exposures acting on each step of the sensitometric strip are 
known, it is possible to estimate by inspection the relative exposures 
required through each of the three tricolor filters to give the same 
result as that obtained by the known exposure through the clear 
filter. This is usually and most conveniently expressed in terms of 
the filter factor for each of the tricolor filters, the filter factor being 
defined as the number by which the exposure incident upon the clear 
filter must be multiplied in order to obtain the value of the exposure 
which must be incident upon the filter in question so that the photo- 
graphic effect on the material exposed through that filter shall be 
equal to that resulting from the exposure through the clear filter. 
In the case illustrated, the exposure increases by consecutive powers 
of 2 from step to step ; that is, it doubles for each successive step of 
the sensitometric scale. The estimation of the filter factors from the 
sensitometric exposure made in this manner will depend to some 
extent on whether the equilization of density be made in the region 
of extremely low densities, in the region of the half-tones, or in the 
region of high densities. This, of course, is due to the gamma wave- 
length effect which, while not great in this particular case, is sufficient 
to affect the values of the estimated filter factors. While it may not 
be possible in the half-tone reproduction of this tricolor exposure to 
detect the small differences that were present in the original, it is quite 
evident from an inspection of the original that it would be necessary 
to multiply the exposure given through the blue filter by a factor of 
approximately 2 in order to make it balance the exposure through the 
clear filter. Since the exposure through the blue filter is already eight 
times that given through the clear, it follows that the exposures which 
would be required to produce balance between the blue and clear filter 
strips must be in the ratio of approximately 1 to 16. The filter factor 
for the tricolor blue filter, therefore, is 16. Likewise, judging the 
green and red exposures, also in the low density (shadow) regions, 
filter factor values of 8 and 6, respectively, are obtained. Now, if 
equilization be transferred to the half-tone regions, a somewhat 
different result is found. For instance, the factors for the blue, green, 



350 



LOYD A. JONES 



[J. S. M. P. E. 



and red filters are 32, 6, and 4, respectively. Equilizing the response 
in the high density (highlight) region, values of 40, 6, and 3 for the 
blue, green, and red are obtained. It is customary to make this 
estimation of filter factor for the half-tone region, thus balancing up 
to a certain extent the gamma-wavelength effect. 

Using this method three numbers are obtained which are a fair 
index of the spectral sensitivity of the material. It is possible, of 
course, actually to read the densities resulting from these tablet 
exposures and to plot the usual density-log E characteristic curves for 




FIG. 67. 



O.Z 

LO& EXPOSURE: 

Density-log E curves for the sensitometric strips illustrated in Fig. 
66(6). 



the four filters. This has been done for the particular sample of film 
from which the illustration in Fig. 66 was made. The curves obtained 
are shown in Fig. 67. In plotting these curves account has been 
taken of the fact that the A, B, and C4 filters had exposures eight 
times as long as those given through the clear filter. The curves are, 
therefore, placed correctly in relation to the log exposure scale. Now 
it is possible to determine by measurement the filter factors in terms 
of the inertia values, or, by drawing a horizontal line through the 
region of half-tones (D = 1.00 is used) and reading off the log E values 
where this horizontal line cuts the four characteristic curves, the 



Mar., 1932] PHOTOGRAPHIC SENSITOMETRY 351 

evaluation of filter factor can be made in terms of the half-tone 
region. Results actually obtained in this manner are as follows: 

D = i.o 



Filter 


Log E 


E 


Factor 


Log E 


E 


Factor 


No. 


2.4 


0.0025 


. . 


T.66 


0.0457 




No. 49 


1.5 


0.0316 


12.5 


1.04 


1.097 


24. 


No. 58 


T.3 


0.0200 


8.0 


0.50 


0.316 


6.9 


No. 25 


1.2 


0.0159 


6.3 


0.24 


0.176 


3.8 



It will be noted that these do not check precisely the estimated 
values already given but are of approximately the same order. There 
is little doubt that greater precision can be obtained by reading the 
densities and plotting the curves as illustrated in Fig. 67, but for all 
practical purposes satisfactory values may be obtained by the estima- 
tion process, especially if the observer has had some experience. 

In Fig. 65 the dotted curve, D, is the spectral sensitivity curve of 
the panchromatic material which was used in making the tricolor 
exposure reproduced at (b) in Fig. 66. The tricolor ratio for this 
material, as estimated by use of the densities lying in the half-tone 
region, is 16-8-6, these numbers being, as mentioned previously, the 
multiplying factors for the green, red, and blue filters, respectively. 
This conveys some idea as to the correlation existing between the 
spectral sensitivity of a photographic material, expressed in terms of 
the sensitivity-wavelength function, as derived by the methods of 
monochromatic sensitometry, and the tricolor ratio values, as derived 
by the methods of selective absorption. 

In order to illustrate further the significance of these tricolor ratio 
values and to enable the reader to obtain a somewhat more definite 
correlation of these values with the spectral sensitivity functions, the 
data in Table XVI are given. These are the tricolor ratios obtained 

TABLE XVI 

Tricolor Ratios for Materials Differing in Spectral Sensitivity 

Filter Factors 
Material No. 49 No. 58 No. 25 

Ordinary 4 

Orthochromatic 6 20 

Panchromatic (Type A) 8 14 12 

Panchromatic (Type B) 10 6 10 

Panchromatic (Type C) 16 8 5 

by estimation in the half-tone region of tricolor exposures made on the 
photographic materials of which the spectral sensitivity is shown 



352 LOYD A. JONES [j. s. M. P. E. 

graphically in terms of wedge spectrograms in Fig. 63. For the first 
material which is sensitive only to blue radiation, the filter factors for 
the green and red filters are extremely high and of no practical interest. 
In the case of the second material which is sensitive only to blue and 
green, the filter factor for the red is of no particular interest since it is 
exceedingly great. For the three panchromatic materials the filter 
factors are as shown and a little study of these and the wedge spectro- 
grams in Fig. 63 will show the correlation between the two modes of 
expressing spectral sensitivity. 

Monochromatic Filters. The general method of isolating spectral 
bands by means of selectively absorbing materials, as described in the 
previous section, may be elaborated considerably and thus provide a 
more detailed analysis of spectral sensitivity. For instance, it is 
possible to obtain filters transmitting very much narrower spectral 
bands than those isolated by the tricolor filters already described. The 
rather misleading name of monochromatic filters is sometimes applied 
to filters which transmit relatively narrow spectral bands. It is very 
difficult to obtain filters which transmit spectral bands less than 
50 mju wide and which at the same time have sufficiently high trans- 
missions at the wavelengths of maximum transmission to be of use for 
practical purposes. By exercising great care, however, the visible 
spectrum extending from 400 to 700 mju may be split into five or six 
non-overlapping parts and it is possible in addition to isolate one or 
perhaps two additional sections in the near ultra-violet between 300 
and 400 m/z. Using selectively absorbing filters of this type and 
applying the same general sensitometric procedure which has been 
described under the tricolor method, six or eight numbers may be ob- 
tained, which are of course the multiplying factors for these narrow 
band transmitting filters. It is obvious that this gives a more 
complete analysis of the spectral sensitivity characteristic and from 
these values it is possible to obtain a fairly precise idea of the shape of 
the sensitivity- wavelength function. The value of such a method 
is somewhat doubtful since it lacks the convenience and brevity of the 
tricolor method, by which the result is expressed in terms of three 
values, and, furthermore, lacks the precision and completeness which 
can be obtained by the methods of monochromatic sensitometry. So 
far as the author is aware this method has not been applied to any 
great extent and it would seem wise, if high precision and complete 
data are required, to adopt directly the methods of monochromatic 
sensitometry; while, if the more convenient and simple specification 



Mar., 1932] 



PHOTOGRAPHIC SENSITOMETRY 



353 



in terms of filter factors is adequate for the occasion, the tricolor 
method seems preferable. 

Progressive Cut Filters. One other modification of the selective 
absorption method has been used and advocated by some workers in 
this field and has some merits. This involves the use of a series of 
filters which cut progressively at shorter and shorter wavelength 
values. As mentioned in the previous section, it is very difficult to 
obtain filters which transmit narrow spectral bands efficiently. The 
total transmission of such filters is usually relatively low; hence the 




500 
WAVELENGTH 



TOO 



FIG. 68. Curves 1 to 8, inclusive, spectral transmission curves of a set 
of progressive cut niters. Curve D, spectral sensitivity curve of panchro- 
matic material. 

illumination which can be applied to the surface of the sensitive 
material is low and, therefore, exposure times are relatively long. 
This difficulty can be avoided by adopting the filters of the progressive 
cut type. Such a set is illustrated in Fig. 68. The filter represented 
by curve No. 1 transmits quite freely all radiations of wavelength 
greater than approximately 660 rmz. Exposures made through such a 
filter, therefore, utilize only that portion of the spectral sensitivity 
of the photographic material which lies on the long wavelength side of 
the filter cut. The dotted curve D (Fig. 68) again represents the 
spectral sensitivity curve of a highly panchromatic material, and an 



354 LOYD A. JONES [J. S. M. P. E. 

inspection of the figure will show to what portion of this sensitivity 
any exposure made through filter No. 1 is due. Filter No. 1, of course, 
is a deep red color. The cut of filter No. 2 moves over to approxi- 
mately 610 m/x, that of No. 3 to 590 m/x, No. 4 to 540 m/z, No. 5 to 
500 mju, No. 6 to 450 imx, No. 7 to 390 m/x, No. 8 to 350 imz. Each 
filter, therefore, includes a somewhat greater portion of the spectral 
sensitivity of the photographic material. Exposures made through 
such a set of filters result in a series of sensitometric curves similar to 
those shown in Fig. 67. The integrated sensitivity for each succes- 
sively widened spectral transmission band may be determined as 
previously, either in terms of the exposure corresponding to the inertia 
point or in terms of exposure required to give some constant density 
value, for instance, D = 1.0. By setting up a series of simultaneous 
equations and inserting the sensitivity values derived from the 
individual density-log E curves, a series of numbers can be obtained 
which represent the integrated sensitivity within a computed wave- 
length region, which, of course, depends upon the transmission 
characteristics of the filters. In this way a fairly good approximation 
to the spectral sensitivity curve may be obtained, although it is 
quite impossible to hope to locate all the maxima and minima which 
may actually occur in a function of this type. 

In general, the same criticism applies to this method as to the 
monochromatic filter method. The results are expressed in terms of 
a relatively large number of numerical values and hence lack the 
convenience and brevity of the tricolor filter method. On the other 
hand, the results obtained are not as precise or complete as those 
derivable by means of monochromatic sensitometry. There are cases 
of course where the expense of the equipment required for mono- 
chromatic sensitometry is prohibitive, and in such cases the pro- 
gressive cut filter method may be very desirable since it offers a 
somewhat more complete analysis of spectral sensitivity than can be 
obtained by the tricolor method. 

REFERENCES 

108 ABNEY, W. DEW.: Phot. J., 6 (1882), pp. 136, 154. 

107 ABNEY, W. DEW.: Phot. J., 13 (1888), p. 2. 

108 ABNEY, W. DEW.: Phot. J., 19 (1895), p. 328. 

109 LEIMBACH, G.: Zeit. Wiss. Phot., 7 (1909), p. 157. 

110 LUCKIESH, M., HOLLADAY, L. L., AND TAYLOR, A. H.: J. Frank. Inst., 196 
(1923), p. 495. 

111 OTASHIRO, T.: Bull. Kiryu Tech. College, Japan (Aug., 1923), No. 2. 



Mar., 1932] PHOTOGRAPHIC SfiNSITOMETRY 355 

112 HELMICK, P. S.: J. Opt. Soc. Amer., 6 (1922), p. 998; 9 (1924), p. 521. 

113 HARRISON, G. R.: J. Opt. Soc. Amer., 11 (1925), p. 341. 

114 JONES, L. A., AND SANDVIK, OTTO: /. Opt. Soc. Amer., 12 (1926), p. 401. 

115 MEES, C. E. K., AND WRATTEN, S. H.: Brit. J. Phot., 54 (1907), p. 384; 
Phot. J., 49 (n. s. 33) (1909), p. 235. 

116 CALLIER, A.: Brit. J. Phot., 60 (1913), p. 972. 

117 HANSEN, G.: Z. Physik, 29 (1924), p. 356. 

118 MILLER, O. E.: Rev. Sci. Instr., 3 (1932), p. 30. 

119 ABNEY, W. DEW.: Phot. J., 19 (1895), p. 328. 

120 EDER, J. M.: Phot. Korr. 39, (1903), p. 426. 

121 WALL, E. J.: Brit. J. Phot., 51 (1904), p. 926. 

ERRATUM 

The author regrets having to point out an error in the wording of the paragraph 
of the paper Photographic Sensitometry , Part I, on page 519 of the October, 1931, 
issue of the JOURNAL, beginning "A very ingenious device. . . " to and includ- 
ing "...approximately 1 to 1,000,000 for each wedge." The paragraph in 
question should read as follows: 

"An ingenious method of obtaining directly the characteristic curve of a 
photographic material was suggested by R. Luther 20 in 1910, using a square 
neutral gray wedge behind which the photographic material under test is exposed. 
The resultant negative, developed preferably to high contrast, after having been 
rotated through 90 degrees with respect to its original position, is placed in 
register with the tablet through which the exposure was made so that the lines 
of equal density on the negative are perpendicular to the lines of equal density 
on the tablet. By direct observation of this tablet-negative combination the 
density-log E characteristic may be seen. Or, by making a print, preferably on 
a high contrast material, a permanent record may be obtained." 



STROBOSCOPIC AND SLOW-MOTION MOVING PICTURES 
BY MEANS OF INTERMITTENT LIGHT* 

H. E. EDGERTON** 



Summary. In a paper published in the June issue of the JOURNAL the author 
showed that mercury-arc lamps when excited by quick violent electrical transients 
make a practical source of intermittent light which is very actinic and has a short 
duration of flash. The timing of the flashes is easily controlled. 

In this present paper, further information regarding the duration and the quality 
of the light are givsn. Also improvements upon the mercury-arc tubes are described 
which simplify the construction of the light-pulse tube and the electrical circuits. 

Uses of intermittent light for taking motion pictures are described and illustrated 
by examples. There are in general two methods of using the intermittent light. 
One method is used to take pictures where the light is caused to flash for each frame 
and the film runs at a continuous speed. The second is used to take stroboscopic 
moving pictures of rapidly moving objects by causing the light frequency to approach 
the frequency of the motion of the object. Examples of the later method are shown, 
these being stroboscopic motion pictures of a crude motion picture claw mechanism 
operating at 30 fps. and of the surges in the valve springs of a gasoline engine 
running at 1930 rpm. 

In a paper published in the June issue of the JOURNAL, the author 
discussed briefly some of the possibilities of the use of intermittent 
light in taking motion pictures and described how the mercury-arc 
tube could be used to produce intense light of short duration. It 
was shown how the method was used to study the angular swinging of 
synchronous machines. Since then considerable improvement has 
been made in the source of intermittent light and several uses for the 
application of the light have been made which are of interest to motion 
picture engineers. 

Before discussing the applications of the intermittent light to 
motion pictures, some data will be given regarding the mercury-arc 
tube as a source of intermittent light. 

The circuit for producing intermittent light from mercury-arc 
tubes has been modified so that nearly any type of mercury-arc lamp 

* Presented at the Fall, 1931, Meeting at Swampscott, Mess. 
** Massachusetts Institute of Technology, Cambridge, Massachusetts. 
356 



STROBOSCOPIC PICTURES 



357 



can be used. This circuit eliminates the necessity for the holding-arc 
electrodes, the auxiliary anodes, the grid around the main anode, and 
the mercury condensation chamber, all of which were required for 
tubes to be used with the circuit which was described before. Tubes 
now need an anode at one end, a small pool of mercury at the other, 
and an external electrode around the mercury pool. The shape of the 
tube may have practically any form. A very convenient shape to use 
is a long slender tube which may be placed at the focus of a parabolic 
reflector in order to concentrate the light. 

The necessary elements and arrangement of a variable frequency 
source of intermittent light are shown in Fig. 1. A source of d-c. 
power is connected to the light-pulse tube through a choke and a 
resistor which are large enough to hold back the current until the tube 
has de-ionized, and still are small enough to allow the condenser to 




FIG. 1 . Elementary wiring diagram showing an arrange- 
ment to produce intermittent light from a mercury-arc 
light-pulse tube. 



charge in time for the next flash. The condenser is connected in 
parallel with the light-pulse lamp, and before a flash the condenser is 
charged so that the anode is positive. The tube is started by applying 
a sudden high voltage to the external connection around the mercury 
pool. This is conveniently accomplished by using a step-up trans- 
former through which a small condenser is discharged by means of a 
switch or a small thyratron. 

TIME DURATION OF THE FLASH OF LIGHT 

One of the most important qualities of the light is the quickness of 
its flash. From a practical consideration the exact time of discharge 
is not of importance except that it must be less than a certain mini- 
mum. This minimum allowable time of flash depends upon the 
specific use which is being made of the tube. 



358 H. E. EDGERTON [j. S. M. p. E. 

If photographs of a moving object are being taken, the minimum 
allowable time of flash must be such that there is no appreciable 
motion. If motion -picture photographs are being taken or projected 
by means of intermittent light, the minimum allowable time of flash 
must be such that there is no appreciable blur on the film or screen. 

Many factors influence the time of flash. Before enumerating 
these a brief discussion of the electrical transients, which are the source 
of the stroboscopic light, will be given. A condenser is charged to a 
certain voltage and then is discharged at the desired instant through 
the mercury-arc tube. This discharge is quite violent and quick and 
causes a pulse of light to be emitted from the tube. 

When a condenser is discharged through a constant resistance the 
current rises to a value determined by the voltage across the con- 
denser divided by the resistance. The current then decreases 
exponentially at a rate determined by the time-constant of the circuit, 
which is smallest for a small resistance. The mercury-arc tube acts 
somewhat like a small resistance and thus the condenser is quickly 
discharged. The light is some function of the current and thus of 
time. Another influencing factor in addition to the resistance of 
the tube and the leads is the inductance of the leads, and this tends to 
make the discharge oscillatory, helping to de-ionize the mercury-arc 
tube so that it will not conduct while the condenser is being charged 
for the next flash. 

Factors that influence the time of discharge and which are believed 
to tend to increase the time are: 

1. Resistance of the leads to the tube. 

2. Inductance of the leads to the tube. 

3. High temperature of the tube. 

4. Long tube dimensions. 

5. Large discharge capacity. . 

6. Low voltage on the condenser. 

The first two do not contribute very much to the time of discharge 
for the usual arrangement. The third temperature may increase 
the discharge time materially but, on the other hand, a hot tube gives 
out much more light than a cold one for the same amount of electrical 
energy input to the circuit. The exact effect of tube dimensions for 
such transients has not been investigated thoroughly to the author's 
knowledge. The remaining two factors capacity and voltage are 
somewhat related since the energy of a flash is proportional to the 
energy stored in the condenser, that is, E 2 C/2 joules. A large capacity 



Mar., 1932] STROBOSCOPIC PICTURES 359 

requires a longer time to be discharged in a circuit of linear resistance, 
and it is believed that somewhat the same phenomena are involved 
with the mercury-arc tube. A high voltage tends to speed up the 
discharge especially since it aids in ionizing the gas in the tube so that 
it will start more quickly. 

A revolving drum camera, built by Mr. C. S. Draper of the 
Aeronautics Department of the Massachusetts Institute of Tech- 
nology, was used for the following experiments to determine the time 
of flash. A drum one foot in diameter was rotated at a speed of 
1800 rpm. The stroboscope tube was covered, except for a narrow 
slit, by a piece of cardboard. The linear velocity of the periphery of 
the drum was 1800/60 X 12 X TT = 1130 inches per second; or one 
inch = 1/1130 sec. = 0.00088 sec. 

The data from several runs are tabulated in Table I. These data 
are approximate but do show the influence of temperature. 

TABLE I 

Tabulation of Data Giving the Time Duration of the Light Flashes 

Length of 63% Time Constant 

Temperature of Capacity in Condenser of Exposure on of Flash in 

Tube Estimated Microfarads Voltage Film in Inches Microseconds 

125 C 2 1350 0.06 53 

40 C 2 1350 less than 0.01 less than 9 

75 C 4 1350 0.02 18 

Tube dimensions: 2 ft. long, 20 mm. in diameter. Leads from condenser to 
tube consist of 12 ft. of lamp cord. 

QUALITY OF THE LIGHT FROM THE TUBE 

The spectrum of the light from a light-pulse tube is radically 
different from the spectrum of the light from the same tube excited 
with direct current. The violent electrical discharge excites many of 
the enhanced or spark lines in the spectrum. As a result there are ten 
or so additional lines in the red and a great many additional ones in 
the green besides other lines. The appearance of the light to the eye 
is yellow- white, which is quite pleasing when compared to the ghastly 
blue of the ordinary mercury-arc lamp. 

The two spectrograms shown in Fig. 2 were taken by Mr. W. E. 
Albertson through the courtesy of Professor G. R. Harrison of the 
Physics Department of the Massachusetts Institute of Technology. 
These photographs show the spectrum of an ordinary mercury-arc 
lamp (lower) and the spectrum of the intermittent mercury-arc tube 



360 



H. E. EDGERTON 



[J. S. M. P. E. 



(above). The additional lines may be observed by comparing these 
two spectrograms. The exposures of these two spectrograms was 



Red Green 



Blue Violet 



Stroboscope arc Nonex tube 




A 5461 4359 3653 3132 2804 

D-C. arc in quartz tube 

FIG. 2. Spectrum of the light from a quartz mercury-arc lamp and from 
a light-pulse stroboscope tube of nanex glass. 

made so that the main arc lines of the two would have approximately 
the same intensity. 

MOTION PICTURES WITH INTERMITTENT LIGHT 

In general, there are two methods of taking motion pictures with 
intermittent light. One method is to synchronize the light with the 
position of the film so that the exposure is properly placed on continu- 
ously moving film, or so that one flash of light will occur when the 
shutter is open for the ordinary method of taking pictures. The 
second method is to synchronize the light with some rotating or 
vibrating object which is to be photographed. Exposures are then 
obtained by random coincidence of a flash of light and an open 
shutter, it being possible to get both more than one exposure on one 
frame or none, depending upon the frequency of the light flashes, the 
exposure angle of the shutter, and the speed of the framing mechanism . 
This will be discussed more completely later. 

Little needs to be said of the first method. For this the intermit- 
tent light is caused to flash at the proper time by the camera mecha- 
nism so that the frames are properly spaced if they are to be projected. 
The flash of light needs to be short enough so that the film does not 
move an appreciable distance while it is on, say, for instance, one 
thousandth of a frame. 

The upper limit of film speed for 16-mm. film with a light whose 
duration is ten microseconds, allowing a motion of one-thousandth 
of a frame, is calculated below: 

The film moves 7.5/1000 mm. in 10 microseconds, whence its velocity 
is (7.5 X 10 6 )/(1000 X 10) = 750 mm. per second, which corresponds 
to 100 frames per second. Allowing the film to move one-hundredth 



Mar., 1932] 



STROBOSCOPIC PICTURES 



361 



of a frame while the light is on increases the maximum allowable 
speed to 1000 frames per second. 

The second method that of synchronizing the light with a moving 
or vibrating object is very useful in taking slow motion pictures 
(stroboscopic) of rapidly moving mechanisms. Say, for instance, it is 
desired to take a moving picture of the claw mechanism of a motion 
picture projector or camera while it is operating at normal speed. 
Obviously, it is impossible to get such a picture without a camera that 
will take at least eight frames while the claw mechanism completes its 
cycle of operations. This calls for a 192-frame-per-second camera if 
the speed of the claw is at 24 frames per second. The pictures would 
not be very clear since the claw would move quite a distance during 
the exposure. 




FIG. 3. Diagram showing the relation between the 
lowest light frequency and the camera speed and shutter 
exposure angle to prevent blank frames. 

The motion of such claw mechanism is easily photographed with a 
motion picture camera if the claw is illuminated with intermittent 
light which has a frequency slightly different from that which cor- 
responds to the speed of the claw mechanism. The claw is seen once 
each revolution in a slightly different position and as a result it ap- 
pears to be moving at a slow speed. This is the well-known strobo- 
scopic effect which has been used for studying moving mechanisms of 
all sorts. The mercury-arc tube is powerful enough to produce 
sufficient light to take motion pictures by this means. 

As has been mentioned previously, the exposure of a film in an 
ordinary motion picture camera does not depend upon the shutter 
angle or speed of framing when intermittent light is used as an 
illumination source. The exposure is entirely determined by the 



362 



H. E. EDGERTON 



[J. S. M. P. E. 



I 

* 

\ 
\ 

JL. 

1 



y 



amount of light in the flash from the tube. There are several possi- 
bilities which must be kept in mind regarding film speed and light 
frequency when taking stroboscopic motion pictures. The ideal 
method is to control the speed of the camera so that only one exposure 
occurs on each frame, but this is not possible with the constant-speed, 

spring-driven cameras. The exposure 
ratio between one and two flashes is not 
objectionable in a projected picture, but a 
blank frame causes a flicker which disturbs 
the continuity of the events. Fig. 3 shows 
the relation that exists between the light 
frequency and that of the camera mecha- 
nism to prevent blank frames, this relation 
being that the lowest frequency of the 
light should be equal to or greater than 
the frequency of the camera (frames per 
second) multiplied by the percentage of 
time that the shutter is open. This limit- 
ing condition spaces two light flashes so 
that if one occurs when the shutter has just 
opened, the other will occur when the 
shutter has just closed. 

Since the shutter does not open and 
close instantaneously and because the light 
is practically instantaneous, it is possible 
to get an incomplete picture. Frame No. 
12 in Fig. 4 is one of these. 

EXAMPLES OF STROBOSCOPIC MOTION PICTURES 
(STROBOGRAMS) 

The motion of a crude 35-mm. claw 
mechanism was photographed with a 16- 
mm. cine kodak, using intermittent light 
which was of a slightly different frequency 
than the claw mechanism. The moving 

pictures were taken with an //1. 9 lens on the standard film at 16 
frames per second. The claw mechanism which was photographed 
was operating at about 30 frames per second. Fig. 4 shows an en- 
larged section of this film. 

The first three frames of Fig. 4 show the claw as it pulls the film 
down. For the frame numbered 4, the mechanism has started to pull 



f 



a? 



2? 



FIG. 4. Stroboscopic 
motion pictures of a claw 
mechanism operating at 30 
frames per second, taken 
with a 16 fps. camera. 



Mar., 1932J 



STROBOSCOPIC PICTURES 



363 



away from the film. Frames numbered 5 to 12 show this drawing 
back in its various stages. The lag of the spring due to its inertia is 
easily observed. The next frames (13 to 21) show the return stroke 
of the claw. The inertia of the spring here causes it to be bent back 
the other way. Frame 20 shows the spring just after it has touched 
the guide and the end of it has bounced back. The remaining frames 
complete the cycle of events, showing the claw as it pulls the film 
down. 

Twenty-four frames are shown in Fig. 4 of a phenomenon that 




FIG. 5. Photograph of the stroboscope 
arranged to take motion pictures of the 
surges of the valve springs of an experi- 
mental engine. 

occurs in one-thirtieth of a second, and so the apparent speed of the 
camera is 24 X 30 or 720 frames per second. This stroboscopic 
method is good only for mechanisms which are periodic, but it is very 
useful for this purpose. 

As a second example, the oscillations or surges of a pair of valve 
springs were photographed in the Aeronautical Power Laboratory at 
the Massachusetts Institute of Technology with the cooperation of 
Mr. C. S. Draper and Mr. Towner. The stroboscope with its para- 
bolic reflector is shown in Fig. 5 together with the experimental 
engine whose valve mechanism was photographed. The pictures 



364 H. E, EDGERTON 

show the rocker arm slowly going up and down, followed by compres- 
sion waves traveling back and forth through the spring. Three 
enlarged pictures which were selected from a 35-mm. motion picture 
film are shown in Fig. 6. The top picture of the left spring shows 
the coils open at the top and compressed at the bottom. The 




FIG. 6. Three enlarged strobo 
scopic photographs from a 35-mm. 
film of surges in a valve spring. 

middle picture shows the spring still opened more between the top 
coils than the bottom. The lower picture shows the spring coils 
widely separated at the bottom and compressed at the top. The 
time of exposure for each of these pictures was about 0.00001 
second. 



SOUND IN THE LOS ANGELES THEATER LOS ANGELES, 

CALIF.* 



D. M. COLE** 

Summary. The sound reproducing equipment used in the Los Angeles Theater 
is described in a general manner. Many refinements have been used in this in- 
stallation, including aids for the hard of hearing, broadcast pick-up, and a public 
address system, which enable the exhibitor to furnish better entertainment and more 
comfort to the patrons. Means are provided for reproducing the picture and the 
accompanying sound in the lounge, and provision is also made for disk reproduc- 
tion, in addition to film reproduction. A reproducer set is also provided for the 
reproduction of non-synchronous commercial records, making possible the running 
of continuous programs for entrance music, exit music, and sound effects. 

The trend in modern theater construction is toward larger and 
better equipped theaters. Mechanical and electrical devices, which 
enable the exhibitor to furnish better entertainment and more comfort 
to patrons, are being used increasingly in new theaters, refinements 
being added as they become available. 

The Los Angeles Theater is an example which included in its 
construction and furnishings all available refinements. The acoustic 
properties of the theater were given careful consideration and, hand in 
hand with good sound equipment, excellent results are being achieved. 
In addition to the sound picture equipment, various attachments and 
special features have been provided. The sound facilities include 
sound picture reproduction, both film and disk for three projectors, 
hard-of -hearing aids, non-synchronous attachment, broadcast pick-up, 
and public address systems. Fig. 1 is a view of the equipment installed 
in the projection room. The amplifiers and control panels are mounted 
on five racks, centralizing all the panels, with the exception of the 
public address control equipment, which is located in a room adjacent 
to the projection room. Two sets of amplifiers are provided, permit- 
ting simultaneous reproduction of two programs; i. e., while sound 
pictures are being shown in the theater auditorium, announcements 

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. 
** Electrical Research Products, Inc., New York, N. Y. 

365 



366 



D. M. COLE 



[J. S. M. P. E. 



can be made to other parts of the theater, if required. The duplicate 
set of amplifiers insures sound in the theater auditorium at all times. 
Switches have been used throughout in this installation, with the 
exception of the inputs connecting the microphones to the mixing 
panel, where jacks are used. Monitoring facilities for both systems 
are provided. Loud speakers of various types to fit the particular 
purpose are installed about the theater to care for the distribution of 
programs. 

Sound Picture Equipment. The sound picture equipment is of 
the largest type of Western Electric equipment supplied for de luxe 




FIG. 1. View of equipment installed in projection booth. 

theaters. The amplifier equipment consists of a voltage amplifier, a 
medium power amplifier, and two high power amplifiers. The ampli- 
fiers, with the exception of the voltage amplifiers, have "built-in" 
rectifiers and filters which furnish plate supply from alternating 
current. The plate current of the voltage amplifiers is obtained 
from the rectifier of the medium power amplifier with which it is 
associated. The filament supply for the medium and high power 
amplifiers is obtained from 110 volts a-c. stepped down to the proper 
voltage. The filament supply for the voltage amplifier is obtained 
from a motor generator set. Horn control panels are provided for 
impedance matching and testing of the horn receivers. Pick-up 



Mar., 1932] 



SOUND IN Los ANGELES THEATER 



367 



equipment is provided to permit the reproduction of either film or 
disk records on any one of three projectors. This equipment is of 
the universal base type. 

Three shallow type stage horns, each equipped with two receivers 
are used behind the screen for obtaining correct illusion and distribu- 
tion of sound. A large sound screen 60 by 40 feet, having a good 




FIG. 2. View of the miniature screen which enables patrons 
to see in the lounge the picture simultaneously projected in the 
theater auditorium. 

frequency transmission characteristic and good light reflecting quali- 
ties, is installed. 

The volume of sound is normally controlled in the projection room 
but an auxiliary fader is available for use in various locations in the 
auditorium. The auxiliary fader is used for previews and premiere 
openings where special attention to volume is essential. 

In the Grand Salon a miniature screen is provided which enables 
patrons to view the picture which is being shown simultaneously in 



368 D. M. COLE [j. s. M. P. E. 

the theater auditorium. This is shown in Fig. 2. The accompany- 
ing sound is reproduced by a loud speaker which is located above 
the screen behind the grille work. 

Loud speakers are provided in two "cry rooms," enabling those 
viewing the picture from this point to hear the accompanying sound. 

Hard-oj f -Hearing Aids. Hard-of -hearing aids enable partially deaf 
patrons to hear both the sound picture reproduction and stage pro- 
grams. Single receivers, provided with head bands, are employed. 
A regulating device in the cord permits the patron to adjust the 
volume to suit his need. The cords are equipped with plugs which 
are plugged into receptables installed on the arms of the seats. An 
a-c. operated amplifier, which obtains a small speech input voltage 
from one of the system amplifiers, furnishes the power for these 
receivers and precludes the possibility of short circuits in the hard- of - 
hearing aid attachment from interfering with the operation of the 
system with which it is associated. 

N on- Synchronous Attachment. For the reproduction of incidental 
music recorded at 78 rpm., a reproducer set is installed in the projec- 
tion room. Two turntables with a fader make possible the running 
of continuous programs for entrance music, exit music, and sound 
effects. 

Radio Broadcasting Feature. Two amplifiers are provided to 
furnish programs over telephone lines to radio broadcasting stations. 
Programs from any of the microphone pick-up points, including the 
broadcasting studio, can be transmitted. The amplifiers are all a-c. 
operated and the necessary impedance matching and isolating trans- 
formers are provided. 

Public Address. The public address portion of this installation 
consists of high quality microphones of the condenser type, with their 
associated amplifiers, control equipment, voltage and power ampli- 
fiers, switching panels, and loud speakers of various types. Micro- 
phone outlets are provided for pick-up from the footlights, stage, the 
orchestra pit, broadcasting studio, foyer, check-room, and lobby. 
Provision is also made for a hanging type microphone over the 
orchestra pit. Suitable mountings are provided, depending on the 
location in which the microphones are used and the function which 
they perform. The microphones are of the same type as those used 
in field and studio recording. A 200-volt dry battery is provided to 
furnish the polarizing voltage for the condenser microphones and the 
plate supply for their associated amplifiers. The low voltage supply 



Mar., 1932] SOUND IN LOS ANGELES THEATER 369 

for filament currents and grid potentials is obtained from the filtered 
output of a motor generator set. The amplifier associated with each 
condenser microphone is so constructed that it is not disturbed by 
shocks, this being accomplished by means of spring suspension 
construction. The microphone pick-up control panel is located in a 
room adjacent to the motion picture room. From this point, the 
operator can observe the results of amplifying speech or music in the 
auditorium. The mixing facilities enable the operator of the public 
address equipment to blend the output of any three microphones, as 
required. Standard studio equipment is provided for this purpose. 
The public address amplifying equipment consists of two voltage 
amplifiers, a medium power amplifier, and two high power amplifiers. 
It should be noted here that additional voltage amplification over that 
needed for sound picture reproduction is required for public address 
work. The power required to operate the public address amplifiers is 
obtained in the same way that the power for the sound picture 
amplifiers is obtained. The medium power amplifier is capable of 
furnishing the plate supply for two voltage amplifiers. 

For general reinforcement work in the theater, large horns equipped 
with high quality receivers, are located over the proscenium arch and 
in the right and left organ grilles. During operation, the volume is 
maintained at a point which creates the illusion to the patron that 
the reinforced sound is coming from the real source. The relation of 
the horns to the pick-up source is very important, and in general it is 
essential that the horns be located directly over and a little forward of 
this point. The dynamic loud speakers, installed in the "cry rooms," 
the Grand Salon, the Main Lounge, and the foyer, furnish incidental 
music to patrons entering and leaving the theater and to those waiting 
about for one reason or another. 

Power Supply. The low voltage power supply for the entire 
installation is obtained from two motor generator sets with associated 
filters. The motor generator sets can be used interchangeably and, in 
emergency, either would handle any load which might be required to 
keep the show running. They furnish low voltage to the condenser 
microphone amplifiers, the voltage amplifiers, the film reproducer 
amplifiers, signal circuits, and the fields of the horn receivers. The 
remainder of the equipment, including the power amplifiers, operates 
from the standard power supply. A voltage control cabinet is pro- 
vided to care for fluctuations in the line voltage. 

Conclusion. All the equipment is of the very highest quality, 



370 D. M. COLE 

from a mechanical and voice and music transmission standpoint. 
With the service rendered by the supplier of the equipment and 
the excellent work of the theater personnel, the system has been kept 
in operation with a minimum of trouble, in spite of the fact that the 
theater operates during long hours. Close cooperation between the 
theater management and the manufacturer of the equipment insures 
maximum use of the equipment, particularly the public address and 
special features. 

There are indications that the larger first-class motion picture and 
legitimate theaters will soon be equipped with facilities similar to 
those enjoyed by the patrons of the Los Angeles Theater. 



THE REDUCING ACTION OF FIXING BATHS ON THE 
SILVER IMAGE* 

H. D. RUSSELL AND J. I. CRABTREE** 



Summary. The extent of the reducing effect of fixing baths on the silver image 
during the progress of fixation is greater than has been generally supposed. For 
example, in sensitometric work it is dangerous to prolong the fixation of motion picture 
positive film in the average fresh potassium alum fixing bath beyond 5 minutes at 
65 F. and with certain highly acid chrome alum baths a measurable degree of re- 
duction occurs even in this short space of time. 

Since little or no reduction of the image occurs in an alkaline hypo solution, 
sensitometric tests should be checked against images fixed in a 25 per cent solution 
of hypo containing 1 per cent of sodium carbonate (anhydrous). 

In regular laboratory work the degree of reduction which takes place in the normal 
time for fixation is usually of no practical importance with the baths in common 
use. In any given bath the rate of reduction increases with the acidity, the tempera- 
ture of the bath, and degree of agitation of the film. 

During use, the reducing action of a fixing bath falls off because it becomes more 
alkaline and accumulates silver thiosulfate which tends to retard the reduction. 

In order to insure the minimum degree of reduction, baths having a minimum 
degree of acidity should be used although such baths have a short life and often do 
not harden satisfactorily. It is therefore necessary to revive such baths either by 
adding further quantities of acid or hardening solution at intervals during use, 
otherwise if the film is not rinsed in water before fixing an objectionable sludge will 
form in the fixing bath. 

The nature of the reduction with the negative emulsions tested was found to be 
almost strictly proportional, and some of the more active baths enumerated could 
therefore be used advantageously for reducing the contrast of photographic images. 

OUTLINE 

I. Experimental methods. 

II. Degree of reduction of the silver image in various fixing baths. 

III. Effect of reduction on shape of characteristic curve. 

IV. Factors affecting the rate of reduction. 

(A) Composition of fixing bath. 

1. Acidity (/>H) of bath. 

2. Sulfite concentration. 

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. Communica- 
tion No. 484 from the Kodak Research Laboratories. 

** Kodak Research Laboratories, Eastman Kodak Co., Rochester, N. Y. 

371 



372 H. D. RUSSELL AND J. I. CRABTREE [j. s. M. p. E. 

3. Hypo concentration. 

4. Hardener concentration. 

5. Nature of hardening agent. 

(B) Temperature of bath. 

(C) Degree of agitation. 

(D) Age of bath before use. 

() Nature of developer and degree of development of image. 

(F) Concentration of exhaustion products. 

(G) Concentration of various addition agents. 
(IT) The presence of oxygen and oxidizing agents. 

V. Factors affecting the rate of reduction in solutions of plain hypo. 

VI. Theoretical discussion. 

VII. Summary. 

VIII. Practical recommendations. 

Although it is well known that under certain conditions a fixing 
bath may exert an appreciable reducing action on the silver image of 
negatives and prints, no precise data have been available on the 
magnitude of this effect with present-day motion picture emulsions. 
However, with the widespread application of sensitometry to every 
branch of photography and especially to the photographic recording 
of sound, the question of the extent of this reaction under practical 
conditions is of increasing importance. 

I. EXPERIMENTAL METHODS 

The emulsions tested are tabulated below. 

Emulsion 

Nature of Motion Picture Film Number 
Panchromatic Negative, Type 2 1218 

Supersensitive Panchromatic Negative, Type 2 1217 

Negative 1201 

Duplicating Negative 1505 

Duplicating Negative 1503 

Duplicating Positive 1355 

Positive 1301 

In the majority of the tests only a relative measure of the degree of 
reduction in a stipulated time was obtained when the film to be bathed 
was developed, fixed in the F-2 fixing bath, washed, and dried before 
treatment. In the other tests the film was developed, fixed in the 
bath under test for twice the time required to clear it, and then 
treated for a further period. A separate test strip of the film treated 
for twice the time to clear it was washed, and the density of the dried 



Mar., 1932] REDUCING ACTION OF FIXING BATHS 373 

strip taken as the density before treatment. Film treated in this 
manner is termed "wet film." 

The film was exposed on the Eastman sensitometer, type 2-B, and 
after processing, was bathed in the various solutions, contained in 
250-cc. cylinders, for a given period of time with little or no agitation. 
The positive film was developed in the D-16* formula to a gamma 
between 1.0 and 1.2, and the negative film in the >-76** formula to a 
gamma between 0.6 and 0.7. All the tests were made with fresh fixing 
baths containing 30.0 per cent hypo unless otherwise stated. 

In several experiments in which the film was agitated continuously 
during bathing, the film was pinned to a small drum immersed in the 
solution to be tested and rotated at a peripheral speed of approxi- 
mately 100 feet per minute. 

The progress of the reduction was determined by measuring the 
density removed in a given time from a step having a known density. 
This degree of reduction in a given time was considered as a relative 
measure of the rate of reduction. 

The pH values of the solutions were determined with organic 
indicators in a manner similar to that described in a previous publica- 
tion by the authors. 1 

H. DEGREE OF REDUCTION OF THE SILVER IMAGE IN VARIOUS 
FIXING BATHS 

The degree of reduction of the silver image in the various fixing 
bath formulas published by the Eastman Kodak Company was 
determined with the emulsions listed previously. The constituents 
of the fixing baths are given in Table I in terms of grams or cubic 
centimeters per liter. 

The degree of reduction at 70F. with dried processed film in the 
fixing baths given in Table I is shown in Table II for various times of 
bathing. The results show that in general the rate of reduction in a 





*D-16 


** D-76 


Elon 


. 3 gram 


2 . grams 


Hydroquinone 


6 . grams 


5 . grams 


Sodium sulfite (desiccated) 


40 . grams 


100.0 grams 


Sodium carbonate (desiccated) 


19 . grams 




Borax 


. . . 


2 . grams 


Citric acid 


. 7 gram 




Potassium metabisulfite 


1 . 5 grams 




Potassium bromide 


0.9 gram 




Water to make 


1.0 liter 


1.0 liter 



374 H. D. Ru JSELL AND J. I. CRABTREE [j. s. M. p. E. 

TABLE I 

Constituents of Fixing Baths Used for Determining Degree of Reduction 



Constituents 



F-l 



F-2 



F-14 



F-16 



F-23 



Hypo 300 grams 300 grams 300 grams 300 grams 300 grams 
Sodium sulfite 

(desiccated) 15 grams 3 grams 7. 5 grams 15 grams 17. 5 grams 

Acetic acid (glacial) 13 cc. 5 cc. 13 cc. 
Sulfuric acid 

(concentrated) ... ... ... 2 cc. 2 cc. 

Potassium alum 15 grams 6 grams 15 grams 

Potassium chrome ... ... ... 15 grams 32 grams 

alum 

Water to make 1 liter 1 liter 1 liter 1 liter 1 liter 




FIG. 1. Effect of reduction on the shape of the characteris- 
tic curve. 



Mar., 1932] 



REDUCING ACTION OF FIXING BATHS 



375 



given fixing bath is dependent upon the state of division of the image. 
In the case of images from fine grained emulsions, the rate of reduction 
is much greater than with coarser grained materials such as motion 
picture panchromatic negative film type 2. It is also seen that the 
reducing action is a minimum in the case of certain potassium alum- 
acetic acid baths, while the maximum effect is obtained with chrome 
alum baths containing sulfuric acid. 

III. EFFECT OF REDUCTION ON THE SHAPE OF THE CHARACTERISTIC CURVE 

The effect of the F-23 fixing bath on the shape of the characteristic 
curve is shown in Fig. 1. With motion picture negative film the 
reducing action is almost truly proportional, while with positive film 
the behavior is between that of a cutting and a proportional reducer. 



IV. FACTORS AFFECTING THE RATE OF REDUCTION OF THE SILVER 
IMAGE IN FIXING BATHS 

At the outset it was considered that the rate of reduction might 
depend upon the following factors which were investigated. 



2.0 - 



(DESICCATED) 
WATER. I UATER 




Z.O 4.O 6.O 8.O Z.O 4.O fe.O & O 

FIG. 2. Effect of />H on the degree of reduction in various fixing baths. 



376 H. D. RUSSELL AND J. I. CRABTREE [j. s. M. p. E. 

A. Composition of Fixing Bath. 1. Effect of pH on the Rate of 
Reduction. The effect of the H of the bath on the rate of reduction 
in the F-l and several experimental formulas is shown in Fig. 2. The 
bath numbered F-x was an experimental bath employed in several 
of the tests throughout the paper and contained 30 per cent hypo, 3 
per cent chrome alum, and 3 per cent sodium bisulfite. The pH 
values of the baths were varied by the addition of either sulfuric acid 
or sodium hydroxide. The graphs show that the rate of reduction 
increases rapidly for a given solution as the pH value is decreased 
below a value of 4.0. As the H increases to values greater than 4.0 
the rate decreases more or less rapidly depending upon the composi- 
tion of the fixing bath. 

TABLE II 

The Degree of Reduction of the Silver Image from Different Emulsions in Various 

Fixing Baths 

Density Removed for Different Times of Bathing at 70 F. 
Fixing Bath Original Density 30 Min. 60 Min. 3 Hours 6 Hours 



Positive Film, Emulsion 


1301 




F-2 


1.60 


0.04 


0.20 


0.34 


0.64 


F-16 


1.60 


0.16 


0.36 


0.64 


1.24 


F-23 


1.60 


0.20 


0.46 


1.42 


1.58 


F-l 


1.60 


0.10 


0.24 


0.40 


0.68 


F-14 


1.60 


0.10 


0.30 


0.50 


0.70 


Super sensitive 


Panchromatic Negative Film 


Type 2, Emulsion 


1217 


F-2 


1.36 


0.06 


0.06 


0.16 


0.30 


F-16 


1.36 


0.06 


0.16 


0.24 


0.36 


F-23 


1.36 


0.12 


0.44 


0.60 


0.98 


F-l 


1.36 


. 08 


0.10 


0.20 


0.40 


F-14 


1.36 


0.08 


0.10 


0.22 


0.42 


Panchromatic Negative Film 


Type 2, 


Emulsion 12L8 




F-2 


1.50 


0.06 


0.08 


0.20 


0.30 


F-1Q 


1.50 


0.06 


0.18 


0.30 


0.40 


F-23 


1.50 


0.10 


0.22 


0.70 


1.18 


F-l 


1.50 


0.08 


0.10 


0.16 


0.30 


F-14 


1.50 


0.08 


0.10 


0.20 


0.40 


Negative Film, Emulsion 


1201 




F-2 


1.50 


0.08 


0.10 


0.16 


0.26 


F-16 


1.50 


0.10 


0.30 


0.40 


0.50 


F-23 


1.50 


0.16 


0.30 


0.70 


1.28 


F-l 


1.50 


0.08 


0.10 


0.20 


0.30 


F-14 


1.50 


0.10 


0.18 


0.28 


0.40 



Mar., 1932] REDUCING ACTION OF FIXING BATHS 377 

TABLE II (continued) 

The Degree of Reduction of the Silver Image from Different Emulsions in Various 

Fixing Baths 

Density Removed for Different Times of Bathing at 70 F. 
Fixing Bath Original Density 30 Min. 60 Min. 3 Hours 6 Hours 



Duplicating Negative 


Film, Emulsion 


1505 




F-2 


1.24 


0.04 


0.12 


0.20 


0.42 


F-W 


1.24 


0.08 


0.20 


0.32 


0.96 


F-23 


1.24 


0.10 


0.30 


0.82 


1.14 


F-\ 


1.24 


0.04 


0.14 


0.24 


0.52 


F-14 


1.24 


0.10 


0.18 


0.34 


0.60 


Duplicating Negative 


Film, Emulsion 


1503 




F-2 


1.40 


0.04 


0.12 


0.20 


0.38 


F-W 


1.40 


0.10 


0.20 


0.42 


0.98 


F-23 


1.40 


0.16 


0.34 


0.90 


1.26 


F-l 


1.40 


0.06 


0.10 


0.16 


0.36 


F-14 


1.40 


0.10 


0.16 


0.26 


0.50 


Duplicating Positive 


Film, Emulsion 


1355 




F-2 


1.64 


0.08 


0.14 


0.24 


0.40 


F-W 


1.64 


0.14 


0.24 


0.40 


0.84 


F-23 


1.64 


0.14 


0.34 


0.86 


1.40 


F-l 


1.64 


0.06 


0.14 


0.34 


0.50 


F-14 


1.64 


0.10 


0.24 


0.40 


0.64 



2. Sulfite Concentration. The effect of the concentration of sulfite 
was determined by the addition of increasing quantities of sodium 
bisulfite to a solution containing 30 per cent hypo. The results in 
Fig. 3 are given for two pH values. The value of 4.4 was that of the 
plain solutions, while the value of 3.0 was chosen arbitrarily and was 
obtained by the addition of sulfuric acid. The data indicate that for 
a solution containing 300 grams of hypo per liter an increase in the 
sulfite concentration at a constant />H value increases the rate of 
reduction of the silver image. The rate of reduction was very much 
greater at a pR value of 3.0 than at 4.4. Similar results were ob- 
tained when equivalent quantities of sodium sulfite were substituted 
for sodium bisulfite. 

3. Hypo. Concentration. A decrease in the concentration of hypo 
for a given sulfite concentration and pH value decreased the rate of 
reduction as is shown by experiments 3, 4, 5, and 6 in Table III-A. 
Tests were also made which indicated that with concentrations of 
hypo greater than 30 grams per liter the rate of reduction decreases. 



378 H. D. RUSSELL AND J. I. CRABTREE [j. s. M. P. E. 

TABLE IH-A 

The Effect of Various Reagents on the Degree of Reduction of the Silver Image in 

Fixing Baths 



Sodium Sulfuric 
Bisulfite Acid Time of 
Nature Hypo (Grams 10% Bathing 
of (Grams per (Cc. per at 70 F. 
No. Bath per Liter) Liter) Liter) (Hrs.) 



Density Color 
Original Re- of 

Density moved Image 



Remarks 



1 


300 




50 


0.0 


1 


.0 


4.4 


3.08 


0.45 


Black 




2 


300 




50 


40.0 


1 


.0 


3.2 


3.08 


1.94 


Brown 




3 


300 




10 


0.0 


1 


.0 


4.4 


3.08 


0.17 


Black 




4 


300 




10 


40.0 


1 


.0 


3.2 


3.08 


1.20 


Brown 




5 


10. 





10 


0.0 


1 


.0 


4.4 


3.08 


0.10 


Black 




6 


10. 





10 


40.0 


1 


.0 


<3.0 


3.08 


0.12 


Black 


Sulfurized 


7* 


300 




50 


40.0 


1 


.0 


3.2 


3.06 


2.00 


Brown 




8* 


75.0 


12.5 


40.0 


1 





3.2 


3.06 


1.40 


Black 




9* 


37. 


5 


6.25 


20.0 


1 





3.0 


3.06 


0.70 


Black 




10* 


18. 


75 


3.12 


20.0 


1 


,0 


3.0 


3.06 


0.46 


Brown 




11 F-l 










1 


,0 


3.8 


3.08 


0.16 






12 F-l 








10.0 


1 


.0 


3.6 


3.08 


0.20 






13 F-l 








20.0 


1. 





3.4 


3.08 


0.35 






14 F-l 


. 




. 


40.0 


1. 





3.0 


3.08 


1.58 




Sulfurized 



Equal ratio of sulfite and hypo. 



3.0 



z.c. 



1.4 



\.TL 



1.0 
Oft 



OA 



POSITIVE: ni_Ni 
ORV 

NO A.G \TACT\OM 




BISULPHITE. 



FIG. 3. Effect of sulfite concentration on the degree of reduction. 



Mar., 1932] 



REDUCING ACTION OF FIXING BATHS 



379 



TABLE III-B 

The Effect of Various Reagents on the Degree of Reduction of the Silver Image in 

Fixing Baths 



No. 


Nature 
of Bath 


Cone, of 

Substance 
Added 
Nature of Substance (Grams 
Added per Liter) 


Time of 
Bathing 
at 70F. 
(Hrs.) 


PH 


Original 
Density 


Color 
Density of 
Re moved I mage 


15 


F-X 








1 


.0 


3.0 


3 


.08 


1.94 


Brown 


16 


F-X 


Silver Iodide 


1 


.0 


1 


.0 


3.0 


3 


.08 


1.55 




17 


F-x 


Silver Iodide 


10 


.0 


1 


.0 


3.0 


3 


.08 


0.30 




18 


F-x 


Silver Iodide 


100 


.0 


1 


.0 


3.0 


3 


,08 


0.10 




19 


F-x 


Silver Bromide 


1 





1 


.0 


3.0 


3 


08 


1.90 




20 


F-x 


Silver Bromide 


10, 





1 


.0 


3.0 


3. 


08 


1.20 




21 


F-x 


Silver Bromide 


100. 





1 


.0 


3.0 


3. 


08 


0.10 




22 


F-x 


Potassium Bromide 


1. 





1 


.0 


3.0 


3.08 


2.00 


Brown 


23 


F-x 


Potassium Bromide 


10. 





1 


.0 


3.0 


3. 


08 


2.30 


Brown 


24 


F-x 


Potassium Bromide 


100. 





1 


.0 


3.0 


3. 


08 


2.55 


Brown 


25 


F-x 


Potassium Iodide 


1. 





1 


.0 


3.0 


3. 


08 


1.94 


Brown 


26 


F-x 


Potassium Iodide 


10. 





1 


.0 


3.0 


3. 


08 


2.50 


Brown 


27 


F-x 


Potassium Iodide 


100. 





1 


.0 


3.0 


3. 


08 


3.08 


Brown 


28 


F-x 


Ammonium Chloride 


10. 





1 


.0 


3.0 


3. 


08 


2.08 




29 


F-x 


Ammonium Chloride 


100. 





1 


.0 


3.0 


3. 


08 


2.50 




30 


F-x 


Ammonium Sulf ate 


100. 





1 


.0 


3.0 


3. 


08 


1.00 




31 


F-x 


Sodium Chloride 


100. 





1 


.0 


3.0 


3. 


08 


1.10 




32 


F-x 


Methylene Blue 


10. 





1 


.0 


3.0 


3. 


08 


1.20 





Further tests were made to determine if the rate of reduction was 
dependent upon the concentration or ratio of sulfite to hypo. The 
results of experiments 7, 8, 9, and 10 in Table III-A indicated that the 
rate of reduction decreased as the concentration of hypo and sulfite 
were decreased in equal proportions. 

4. Hardener Concentration. The effect of hardener concentration 
on the rate of reduction, degree of hardening, and H value of the 
F-l* and F-2** formulas is shown in Fig. 4 from which it is seen that, 
if the hardener concentration is decreased to one-half its normal value, 
the degree of reduction is also decreased approximately one-half, 

* F-l. For use add 125 cc. .F-la hardener to 1.0 liter of hypo solution. 
** F-2. For use add 50 cc. F-2a hardener to 1.0 liter of hypo solution. 



Hardener Formulas 



Sodium sulfite (desiccated) 
Acetic acid (glacial) 
Potassium aluminum alum 
Water to make 



F-la 

120 grams 
105 cc. 
120 grams 
1 liter 



F-2a 

60 grams 
100 cc. 
120 grams 
1 liter 



380 



H. D. RUSSELL AND J. I. CRABTREE [j. s. M. P. E. 



while the degree of hardening is not seriously affected. A further 
decrease in the hardener concentration does not produce a correspond- 
ing decrease in the rate of reduction and lowers the degree of harden- 
ing to a value which is too low for practical purposes. 

5. Nature of Hardening Agent. In Table IV are given figures 
comparing the extent of the reduction, obtained with the F-l and F-IQ 
formulas, for equal pH values. The pH values were adjusted by the 
addition of either sodium hydroxide or sulfuric acid. 

The results in Table IV indicate that the rates of reduction were 



0.4 




ZLOCf 

no' 

140* 



5.0 



4.0 



3.0 



F--I FIXING. 
POSITIVE. FIL.M 
10 F-. 

REDUCTION 
WET FIUN'V 
ISO AOTACTIOH 
4.O 



0.4 



o.z 





-bl 



F-Z. F-lXtlHQi BATH 
POSITIVE 



RCDUCTIOK 
W 
MO 

4.0 HOURS 





FIG. 4. Effect of the hardener concentration on the degree of reduction, degree 
of hardening, and />H of the F-l and F-2 formulas. 

similar for potassium alum and chrome alum baths for equal sulfite 
concentrations and pH values. 

B. Effect of Temperature on the Rate of Reduction. The effect of 
temperature on the rate of reduction of the silver image with dry 
motion picture positive film is shown in Fig. 5 for the F-2 and 7^-23 
fixing baths. The results indicate that as the temperature is increased 
from 50 to 90 F., the rate of reduction also increases and the effect of 
temperature was much greater with the F-23 formula than with the 
F-2 formula. In the F-23 bath the degree of reduction in a given 
time was increased approximately ten times for an increase in tempera- 



Mar., 1932] 



REDUCING ACTION OF FIXING BATHS 



381 



TABLE III-C 



The Effect of Various Reagents on the Degree of Reduction of the Silver Image in 

Fixing Baths 



No. 


Cone, of 
Substance Time of 
Nature of Added Bathing 
Nature Treatment Substance (Grams at 70 F. Original 
of Bath with Gas Added per Liter) (Hrs.) pH Density 


Color 
Density of 
Removed Image 


33 


F-X 


None 






2 


.0 


3.0 


a 


.08 


2.30 




34 


F-x 


Air 






2 


.0 


3.0 


3 


.08 


2.54 




35 


F-x 


Carbon 




















Dioxide 


2 


.0 


3.0 


3 


.08 


2.38 




36 


F-l 


None 






2 


.0 


3.8 


3 


.08 


0.80 




37 


F-l 


Air 






2 





3.8 


3 


08 


1.80 




38 


F-l 


Carbon 




















Dioxide 


2. 





3.8 


3. 


08 


1.40 




39 


F-l 








6. 


o 


3.8 


3. 


17 


0.30 


Black 


40 


F-l 




Sodium 
























Perborate 


10 


6. 





3.8 


3. 


17 


0.50 


Black 


41 


F-l 




Hydrogen 


100 cc. 


6. 





3.8 


3. 


17 


0.45 


Black 


42 


F-x, 




Peroxide 




1, 





3.0 


2, 


,10 


0.90 


Brown 


43 


F-x 




Sodium Sulfate 


10 


1. 





3.0 


2. 


10 


0.70 


Brown 


44 


F-x 




Sodium Sulfate 


100 


1. 





3.0 


2. 


10 


0.36 


Black 


45 


F-x 




Sugar 


10 


1. 





3.0 


2. 


10 


0.90 


Brown 


46 


F-x 




Sugar 


100 


1. 





3.0 


2. 


10 


0.70 


Brown 


47 


F-x 




Glycerin 


10 


1. 





3.0 


2 


10 


0.80 


Brown 


48 


F-x 




Glycerin 


200 


1. 





3.0 


2. 


10 


0.50 


Brown 



TABLE IV 

A Comparison of the Degrees of Reduction in Potassium Alum and Chrome Alum 

Fixing Baths 



Concentration 
of Alum 
Fixing (Grams 
Bath per Liter) 


Concentration 
of Sulfite 
(Grams 
per Liter) pH 


* Time of 
Bathing 
(Hours) 


Original 
Density 


Density 
Removed 


F-l 


15 


15 


3.8 


4.0 


1.60 


0.30 


F-l 


15 


15 


3.6 


4.0 


1.60 


0.40 


F-l 


15 


15 


3.4 


4.0 


1.60 


0.94 


F-l 


15 


15 


4.0 


4.0 


1.60 


0.22 


F-l 


15 


15 


4.8 


4.0 


1.60 


0.16 


F-IQ 


15 


15 


3.4 


4.0 


1.60 


0.96 


F-IQ 


15 


15 


3.2 


4.0 


1.60 


1.40 


F-IQ 


15 


15 


3.0 


4.0 


1.60 


1.50 


F-IQ 


15 


15 


4.0 


4.0 


1.60 


0.30 


F-IQ 


15 


15 


4.4 


4.0 


1.60 


0.18 



* Positive film (wet). 



382 H. D. RUSSELL AND J. I. CRABTREE [j. s. M. p. E. 

ture from 65 to 95 F., while with the F-2 formula the increase was 
only about four times. 

C. Effect of Agitation on the Rate of Reduction. -The effect of 
agitation on the rate of reduction with wet and dry film is shown in 
Table V. 

TABLE V 

Effect of Agitation on the Rate of Reduction at 70 F. 



Fixing 
Bath 


Emulsion 


Original 
Density 


Time of 
Bathing 


Constant 
Wet Film 


Density Removed 
Agitation No Agitation 
Dry Film Wet Film Dry Film 


F-2 


1301 


1.70 


30 Min. 


0.14 


0.10 


0.00 


0.00 


F-2 


1301 


1.70 


4Hrs. 


1.40 


1.16 


0.60 


0.30 


Hypo 30% 


1301 


1.70 


1 Hr. 


0.50 


0.60 


0.06 


0.08 


Hypo 30% 


1301 


1.70 


4Hrs. 


1.30 


1.30 


0.16 


0.19 


F-23 


1218 


1.34 


30 Min. 


0.34 


0.18 


0.12 


0.06 


F-23 


1218 


1.34 


1 Hr. 


0.52 


0.36 


0.32 


0.14 



The results indicate that with constant agitation, as compared with 
no agitation, the amount of reduction in the case of positive film was 
increased about ten times, while with negative film the rate was ap- 
proximately doubled. Also, the degree of reduction obtained was 
much greater with the wet film previous to drying than with the dry 
film. 

D. Effect of Age before Use. The effect of age of the fixing baths 
before use on the rate of reduction is shown in Table VI, from which it 
is seen that: (1) the rate of reduction in potassium alum fixing baths 
did not change appreciably on storage for 10 days before use, and 
(2) the reducing action of the chrome alum fixing baths decreased with 
age owing to an increase in the pJI value of the solutions. 

TABLE VI 

Effect of Age before Use on Rate of Reduction 



Fixing 
Bath 


Age 

(70F.) 


PH 


*Time of 
Bathing 


Original 
Density 


Density 
Removed 


F-l 


Fresh 


3.8 


4.0 Hrs. 


1.54 


0.30 


F-l 


10 days 


3.8 


4.0 Hrs. 


1.54 


0.28 


F-2 


Fresh 


3.6 


4.0 Hrs. 


1.54 


0.26 


F-2 


10 days 


3.6 


4.0 Hrs. 


1.54 


0.26 


F-1Q 


Fresh 


3.4 


4.0 Hrs. 


1.54 


0.98 


F-IQ 


10 days 


3.8 


4.0 Hrs. 


1.54 


0.60 


F-23 


Fresh 


3.2 


4.0 Hrs. 


1.54 


1.22 


F-23 


10 days 


3.6 


4.0 Hrs. 


1.54 


0.80 



Positive film (dry). 



Mar., 1932] REDUCING ACTION OF FIXING BATHS 383 

E. Effect of Nature of Developer and Degree of Development. At 
the outset it was considered that the rate of reduction under any 
given conditions would depend on the size of the silver grains which, 
in turn, is determined by (a) the nature of the emulsion, (b) the nature 
of the developer, and (c) the degree of development or "gamma." 

Motion picture panchromatic negative film was developed in 
formulas D-16 and D-76 to equal gammas and then bathed in the 
F-23 fixing bath. From the results in Table VII it is seen that equal 
degrees of reduction were obtained for equal densities regardless of 
the degree of development or the nature of the developer. 

TABLE VII 

Effect of Degree of Development and Nature of Developer on the Degree of Reduction 



Developer 


Fixing 
Bath 


Time of 
Dev. 
(Min.) 


Gamma 


Original 
Density 


Density 
Removed 


* Time of 
Bathing 


ZM6 


7^-23 


2.5 


0.30 


0.60 


0.10 


1.0 Hr. 


D-7Q 


7^-23 


4.0 


0.30 


0.60 


0.08 


1.0 Hr. 


D-1Q 


77-23 


5.0 


0.80 


0.60 


0.10 


1 . Hr. 


D-7Q 


.F-23 


15.0 


0.80 


0.60 


0.11 


1 . Hr. 



* Negative film (emulsion 1218) wet. 

F. Effect of Exhaustion Products on Rate of Reduction. With use 
the chemical nature of the fixing bath changes. The undeveloped 
silver halide grains are dissolved from the emulsion and accumulate 
in the bath as complex silver thiosulf ates and sodium halides. Experi- 
ments 15 to 21, inclusive (Table III-B), indicate that the addition of 
silver bromide or silver iodide to the F-x fixing bath decreases the rate 
of reduction. 

Practical exhaustion tests were made with the F-2 and F-23 formula 
in order to determine the effect of exhaustion with developed and 
undeveloped positive film on the rate of reduction. The results are 
shown in Fig. 6 from which it is seen that the rate of reduction is less 
in a bath exhausted with developed film than in one exhausted with 
undeveloped film. In the case of the bath exhausted with 
undeveloped film, the pH value remained practically constant, while 
with the developed film the pH of the bath gradually increased during 
exhaustion, which may have caused a decrease in the rate of reduction. 

The effect of removing the silver from an exhausted fixing bath on 
the rate of reduction was investigated. The silver was removed by 
an electrolytic method similar to that used in actual practice. 2 The 



384 



H. D. RUSSELL AND J. I. CRABTREE [j. s. M. p. E. 



F-2 fixing bath was exhausted with undeveloped motion picture 
panchromatic negative film (type 2) to the extent of 250 feet per 
gallon when the silver content was 7 grams per liter. The silver was 
then removed by electrolysis and the bath exhausted further to 200 
feet per gallon or a total footage of 450 feet per gallon which is equiva- 
lent to 13 grams of silver per liter. The pH value and sulfite 



O.fc 



14 
I.I 
1.0 
0.8 
O.Q, 
0.4 
O.Z 



-t FIXINO BATH 

HO Afe \TI\T\O IX 
POSITIVE. F\_|v\ 




FIG. 5. Effect of temperature on the degree of reduction 
with the F-2 and .F-23 fixing baths. 

concentration of the solution changed during the electrolysis but 
were maintained constant by additions of sulfite and alkali. The 
reduction tests were made with wet motion picture positive film dur- 
ing the last stage of the electrolysis, that is, when the solution con- 
tained less than 3 grams of silver per liter and also after all the silver 
was removed. The tests in every case indicated that the degree of 
reduction obtained in an exhausted F-2 fixing bath from which the 



Mar., 1932] 



REDUCING ACTION OF FIXING BATHS 



385 



silver had been removed was less than that obtained in the fresh 
solution. 

The fixing bath also becomes contaminated during use with 
partially exhausted developer, which in the case of a hydroquinone 
developer consists of sodium halides, sodium sulfite, hydroquinone 
sulfonates, and alkali. The hydroquinone sulfonates are the result 



UtMOEV/E-UOPEO 
ATM e.XHAAJ*TEO WITH 




FIG. 6. Effect of exhaustion on the degree of reduction 
with the F-2 and 7^-23 fixing baths. 

of the reaction between quinone and sulfite, quinone being an end- 
product of the reduction of the exposed silver halide by hydroquinone. 
The alkali in the developer decreases the acidity of the fixing bath, 
in which case the rate of reduction would also decrease. 

The effect of the addition of an exhausted developer on the rate of 
reduction was tested by the addition of: (1) an oxidized D-IQ de- 



386 H. D. RUSSELL AND J. I. CRABTREE [j. s. M. p. E. 

veloper, and (2) quinone to the F-x fixing bath. The D-IQ developer 
was oxidized by bubbling air through the solution until it would no 
longer develop. Two hundred cubic centimeters of such a developer 
were evaporated by boiling to a volume of 20 cc., and added to 250 
cc. of the fixing bath. A comparison made between this solution and 
one containing 1 per cent quinone for equal pH values indicated that 
these products have very little, if any, effect on the rate of reduction. 

G. Effect of Miscellaneous Addition Agents. Various chemicals 
which are not usually considered as oxidizing agents for silver were 
added to the F-x fixing bath as described below. Potassium bromide 
and potassium iodide were added in concentrations ranging from 0.1 
per cent to 10 per cent (experiments 22-27, Table III-B). Both 
chemicals increased the rate of reduction, the potassium iodide being 
more effective for a given concentration than the potassium bromide. 

Ammonium chloride increased the rate of reduction for concentra- 
tions between 1 per cent and 10 per cent while the addition of 10 per 
cent sodium chloride or 10 per cent ammonium sulfate had little or no 
effect on the reaction (experiments 28-30, Table III-B). Since 
ammonium chloride and ammonium sulfate both tend to decrease the 
clearing time in a fixing bath, further tests were made. The effect of 
these salts on the rate of reduction and clearing times of undeveloped 
positive and negative film is given in Table VIII for the F-2 and F-23 
formulas, from which it is seen that the addition of either ammonium 
chloride or ammonium sulfate increased the clearing time of positive 
film, while in the case of negative film the clearing time was decreased. 
A concentration of either salt between 2.5 per cent and 5.0 per cent 
produced the greatest decrease in the clearing times, the chloride 
being more effective than the sulfate. The above concentrations of 
ammonium chloride also increased the rate of reduction to the greatest 
extent, while an equal quantity of the sulfate did not affect the reac- 
tion. With the 7^-23 formula, the rate of reduction increased up to a 
concentration of 300 grams per liter, but beyond this concentration 
the rate began to decrease. 

This critical point does not correspond with the concentration of 400 
grams per liter of hypo which gives a minimum clearing time with 
motion picture panchromatic negative film. 

The addition of restraining agents such as sodium sulfate, sugar, and 
glycerin decreased the rate of reduction (experiments 43-48, Table 
III-C). Sodium sulfate in this respect was more effective than either 
of the other chemicals for equal concentrations. 



Mar., 1932] 



REDUCING ACTION OF FIXING BATHS 



387 



TABLE VIII 



Effect of Ammonium Chloride and Ammonium Sulfate on the Degree of Reduction 
in the F-2 and F-23 Formulas 



Bath 


Per Cent * Time of 
Ammonium Bathing 
Chloride (Hours) 


Original 
Density 


Density 
Removed 


Time to 
Clear 
Positive 

(Sec.) 


Time to 
Clear 
Negative 
(Sec.) 


F-23 





4 


1.60 


0.76 


35 


240 


F-23 


1.0 


4 


1.60 


0.86 


35 


115 


F-23 


2.5 


4 


1.60 


1.10 


35 


95 


F-23 


5.0 


4 


1.60 


0.96 


40 


85 


F-23 


10.0 


4 


1.60 


0.90 


50 


100 




Per Cent 
Ammonium 
Sulfate 












F-23 





4 


1.60 


0.76 


35 


240 


F-23 


1.0 


4 


1.60 


0.60 


35 


125 


F-23 


2.5 


4 


1.60 


0.70 


40 


105 


F-23 


5.0 


4 


1.60 


0.60 


50 


130 


F-23 


10.0 


4 


1.60 


0.48 


60 


170 




Per Cent 
Hypo 












F-23 


30 


4 


1.60 


0.75 


35 


240 


F-23 


40 


4 


1.60 


0.60 


50 


120 


F-23 


60 


4 


1.60 


0.56 


70 


220 


F-23 


80 


4 


1.60 


0.40 


80 


>300 




Per Cent 
Ammonium 
Chloride 












F-2 





4 


1.60 


0.30 


35 


240 


F-2 


1.0 


4 


1.60 


0.30 


35 


115 


F-2 


2.5 


4 


1.60 


0.40 


35 


95 


F-2 


5.0 


4 


1.60 


0.40 


40 


85 


F-2 


10.0 


4 


1.60 


0.20 


50 


100 




Per Cent 
Ammonium 
Sulfate 












F-2 





4 


1.60 


0.30 


35 


240 


F-2 


1.0 


4 


1.60 


0.30 


35 


125 


F-2 


2.5 


4 


1.60 


0.30 


40 


105 


F-2 


5.0 


4 


1.60 


0.22 


50 


130 


F-2 


10.0 


4 


1.60 


0.10 


60 


170 



* Positive film (dry). 



388 H. D. RUSSELL AND J. I. CRABTREE [j. s. M. P. E. 

H. Effect of Oxygen and Oxidizing Agents. The effect of oxygen 
on the rate of reduction in the F-x and F-l fixing baths is shown in 
Table III-C (experiments 33 to 38, inclusive). In the tests, air and 
carbon dioxide were bubbled through the fixing baths for two hours. 
The results indicated that the effect of bubbling air is probably a 
result of the increased agitation. The rate of reduction, however, 
was slightly greater with air than with carbon dioxide. 

Further tests were made in which wet positive film, which was 
flashed to a uniform density and developed in the D-IQ formula, was 
bathed in the F-2 and F-x formulas. The strips were suspended above 
the baths in such a manner that only part of the film was totally 
immersed. The part above the solution was moistened with the 
solution at 1 -minute intervals throughout the time of bathing. With 
the F-2 formula the density above the solution was decreased to a 
greater degree than that which was immersed in the bath, while with 
the F-x formula the reverse effect was obtained. 

The addition of oxidizing agents such as hydrogen peroxide and 
sodium perborate slightly increased the rate of reduction, as shown by 
experiments 39-41, Table III-C. The effect on the silver image of the 
addition of methylene blue to hypo solutions has been determined 
by one of the authors, 3 who found that under certain conditions 
methylene blue produced reversed dye images. The effect of methy- 
lene blue on the rate of reduction of the silver image was determined 
when added to the F-x formula and the fixing solution recommended in 
the above publication. In each case the low densities of the sensito- 
metric strips were reduced very rapidly, while with the high densities, 
the rate was similar to that of the bath without methylene blue. 
Methylene blue produced the greatest effect when strips were bathed 
in a solution of the dye previous to immersion in the fixing solution. 

To Summarize: From the above tests it was concluded that for a 
given fixing bath containing alum, sulfite, acid, and hypo, (1) the rate 
of reduction increased rapidly if the />H value of the bath was reduced 
below 4.0, (2) for pR values greater than 4.0 the degree of reduction 
was of a much lower order of magnitude, (3) for a given hypo 
concentration and a />H value less than 4.0, an increase in the sulfite con- 
centration increased the rate of reduction, (4) for a given sulfite 
concentration and a />H value less than 4.0 an increase in the hypo con- 
centration up to 300 grams per liter increased the rate of reduction 
but with concentrations of hypo greater than 300 grams per liter, the 
rate of reduction decreased. 



Mar., 1932] REDUCING ACTION OF FIXING BATHS 389 

V. FACTORS WHICH INFLUENCE THE RATE OF REDUCTION IN SOLUTIONS 

OF PLAIN HYPO 

The chemicals and reagents listed in Table III-A, -B, and -C were 
added to a solution containing 300 grams of hypo per liter, but no 
noticeable increase in the rate of reduction was observed. From 
these experiments it was concluded that the nature of the reducing 
action in plain hypo solutions was different from that in acid sulfite 
fixing baths. 

The greatest increase in the rate of reduction was obtained when 
oxygen or air was bubbled through the solution. The effect of 
bubbling various gases through plain hypo solutions is shown in 
Table IX. 

TABLE IX 

Effect of Various Gases on the Rate of Reduction of the Silver Image in Hypo 

Solutions 





Gas 


* Time of 
Bathing 


Original 
Density 


Density 
Removed 


P 

Before 
Treatment 


EE 

After 
Treatment 


1. 


None 


2 


Hrs. 


2. 


10 


0. 


.10 


6. 





6 


.0 


2. 


Air 


2 


Hrs. 


2. 


10 


1. 


08 


6. 





7 


.5 


3. 


Air 


2 


Hrs. 


2. 


10 


0. 


86 


11. 





11 


.0 


4. 


Oxygen 


2 


Hrs. 


2. 


10 


2. 


00 


6.0 


7, 


5 


5. 


Nitrogen 


2 


Hrs. 


2. 


10 


0, 


05 


6. 





6 


.0 


6. 


Carbon Dioxide 


2 


Hrs. 


2. 


10 


0. 


20 


6. 





5 


,2 


7. 


Sulfur Dioxide 


2 


Hrs. 


2. 


10 


0. 


30 


6. 





3 


,4 



* Positive film (dry). 

All the gases were bubbled through 250 cc. of the solution at a rate 
equal to 200 cc. per minute. Throughout the period of bathing 
sulfur dioxide was bubbled through the solution until the hypo 
sulfurized, which required about 15 minutes. The slight increase in 
the rate of reduction with carbon dioxide and sulfur dioxide is possibly 
due to the decrease in pH value. 

Further experiments were made in which the air and other gases 
were removed from a 30 per cent solution of plain hypo by means of a 
vacuum pump. An image on positive film bathed in this solution was 
not reduced in 10 hours, while a density of 0.44 was removed from a 
density of 1.10 in a similar solution from which the air had not been 
removed. 

The above tests indicate that air or oxygen is a very important 
factor in the bleaching action of solutions of plain hypo. It was also 
observed that when air or oxygen is bubbled through a solution of 



390 H. D. RUSSELL AND J. I. CRABTREE [j. s. M. p. E. 

plain hypo in which a silver image is being reduced, the solution be- 
comes more alkaline. The changes in alkalinity take place only in 
the presence of a silver image. When air or oxygen was bubbled 
through a solution of hypo without a silver image, no increase in 
alkalinity occurred. The change in alkalinity probably results from 
the oxidation by oxygen of the finely divided silver to silver oxide, 
which is dissolved by the hypo, forming complex silver thiosulfates 
and sodium oxide. The sodium oxide would exist in such a solution as 
sodium hydroxide, which is very alkaline. 

In experiment 3, Table IX, the hypo solution was made alkaline by 
the addition of sodium hydroxide. A comparison of the rate of 
reduction with that in experiment 2 indicates that the reduction in 
alkaline hypo is less than that in plain hypo. 

The effect of the concentration of hypo on the rate of reduction of 
the silver image in solutions of plain hypo is shown in Table X. 

TABLE x 

Effect of Concentration of Hypo on Rate of Reduction 



Concentration of 
Hypo 
(Grams per Liter) 


Original 
Density 


Density 
18 Hours 


Removed 
36 Hours 


800 


2.06 


0.00 


0.00 


400 


2.06 


0.18 


0.34 


300 


2.06 


0.26 


0.62 


200 


2.06 


0.36 


0.70 


100 


2.06 


0.36 


1.00 


10 


2.06 


0.12 


0.16 


1 


2.06 


0.00 


0.00 



The results indicate that the rate of reduction of the silver image in 
plain hypo solutions increases as the concentration of hypo is de- 
creased from 800 grams per liter to 100, and then decreases for a 
further decrease in the hypo concentration. 

VI. THEORETICAL DISCUSSION 

The chemical reaction involved in the reduction of the photographic 
image in an acid fixing bath is probably one of oxidation of the silver 
to a soluble compound. The reaction may be represented by the 
following equation: 

(7.) 2Ag + H 2 S 2 3 + O ^^ Ag 2 S 2 ? + H 2 O 

Silver Thiosulfuric Oxidizing Silver Water 

Acid Agent Thiosulfate 

The silver thiosulf ate formed readily dissolved in the excess hypo. 



Mar., 1932] REDUCING ACTION OF FIXING BATHS 391 

Although the exact chemical nature of the oxidizing agent is un- 
known, F. Foerster 4 ' 6 ' 6 ' 7 - 8 and his colleagues have shown that addition 
compounds of certain sulfur acids with sulfur dioxide can exist in a 
fixing bath and H. Bassett and R. G. Durrant 9 have suggested that 
these probably react as oxidizing agents. 

Foerster and Vogel 8 have prepared the yellow addition compound 
(K^jOa.SC^) by the action of sulfur dioxide on a potassium thiosulfate 
solution. They claim that the yellow color of an acidified sulfite and 
hypo solution is due to such compounds rather than colloidal sulfur. 

Other yellowish colored addition compounds of sulfur dioxide are 
recorded in the literature such as: 

1. H 2 S 2 3 (S0 2 )x 

2. H 2 S0 3 .(S0 2 )x 

3. HCNS.(SO 2 )x 

4. (HO) 2 S.(S0 2 )x 

5. HI.(SO 2 )x 

The effect of these compounds on the silver image was investigated. 
Various concentrations of the potassium salts were added to a 5 per 
cent solution of sodium sulfite acidified with sulfuric acid. The re- 
sults are given in Table XI from which it is seen that an acidified solu- 
tion of sulfite and iodide reduced the silver image very rapidly, while a 
similar solution containing potassium bromide did not affect the image 
to any great extent. From the standpoint of chemical composition 
bromide forms an addition compound with sulfur dioxide similar to 
that of the iodide, which is colorless. 

The effect of the iodide-sulfur dioxide compound on the silver image 
explains the increase in the rate of reduction obtained when potassium 
iodide was added to the F-x fixing bath. A similar increase, although 
not as great, was obtained when potassium bromide was added to the 
F-x bath, which cannot be explained on the basis of the formation of 
these addition compounds. 

The rate of reduction with the addition of potassium thiocyanate to 
an acidified solution of sulfite was considerably less than that with the 
addition of potassium iodide. The mixture of sulfite and thiocyanate 
without acid attacked the gelatin and removed the emulsion from the 
support. The solution of 5 per cent sodium sulfite with acid did not 
attack the silver image and even a highly concentrated yellow solution 
of metabisulfite (experiment 3) did not reduce the silver image, which 
indicates that this addition compound is not an oxidizing agent for 
silver. 



392 



H. D. RUSSELL AND J. I. CRABTREE [j. b. M. P. E. 



TABLE XI 

The Effect of Addition Compounds with Sulfur Dioxide on the Reduction of the 

Silver Image 



Exp 
No. 


Sub- 
. stance 
Added 


Sodium Sulfuric Time of 
Sulfite Acid Bathing 
Grams (Grams 10% at 
per per (Cc. per 70 F. Original 
Liter Liter) Liter) (Hrs.) H Density 


Color 
Density of 
Removed Solution 


Color 
of 
Image 


1 






50 


200 1 


.0 


3.0 


2.84 







Colorless 


Black 


2 






50 


1 


.0 


9.0 


2.84 







Colorless 


Black 


3 


K 2 S2O 5 


400 




50 1 


.0 


3.0 


2.84 







Yellow 


Black 


4 


KBr 


10 


50 


200 1 


.0 


3.0 


2.84 





.12 


Colorless 


Black 


5 


KBr 


100 


50 


200 1 


.0 


3.0 


2.84 





.14 


Colorless 


Black 


6 


KBr 


10 




4.0 1 


.0 


3.0 


2.84 





.0 


Colorless 


Black 


7 


KBr 


10 


50 


1 


.0 


9.0 


2.84 





.0 


Colorless 


Black 


8 


KBr 


100 


50 


1 


.0 


9.0 


2.84 





.0 


Colorless 


Black 


9 


KI 


10 


50 


200 1 


.0 


3.0 


2.84 


2 


.50 


Yellow 


Yellow 


10 


KI 


100 


50 


200 1 


.0 


3.0 


2.84 


2.74 


Yellow 


Yellow 


11 


KI 


10 


50 


1 


.0 


9.0 


2.84 





.36 


Colorless 


Black 


12 


KI 


100 


50 


1 


.0 


9.0 


2.84 





.02 


Colorless 


Black 


13 


KI 


10 




4.0 1 


.0 


3.0 


2.84 





.00 


Colorless 


Black 


14 


KI 


100 


. . 


4.0 1 


.0 


3.0 


2.84 


Gelatin was removed 


15 


Na 2 S2O 4 


250 


50 


200 1 


.0 




2.84 


1 


.42 


Yellow 


Brown 


16 


Hypo 


300 


50 


200 1 


.0 


3.0 


2.84 


1 


.70 


Yellow 


Brown 


17 


Hypo 


300 


50 


1 


.0 


9.0 


2.84 





.20 


Colorless 


Black 


18 


Hypo 


100 


50 


200 1 


.0 


3.0 


2.84 


1 


.32 


Yellow 


Brown 


19 


KCNS 


10 


50 


200 1 


.0 


3.0 


2.84 


0.00 


Yellow 


Black 


20 


KCNS 


100 


50 


200 1 


.0 


3.0 


2.84 





.30 


Yellow 


Black 


21 


KCNS 


100 


50 


1 


.0 


9.0 


2.84 


Gelatin was removed 


22 


KCNS 


100 




4.0 1 


.0 


3.0 


2.84 


Gelatin was removed 


K 2 S2Os = Potassium 


Metabisulfite 




KBr 


= Potassium Bromide 




KI 


= Potassium 


Iodide 
















Na 2 S 2 O4 = Sodium Hydrosulfite 




KCNS 


= Potassium 


Thiocyanate 



Sodium hydrosulfite forms a permanent yellow solution when acidi- 
fied in the presence of sulfite, and this is due to a thiosulfate-sulfur 
dioxide complex, according to Bassett and Durrant. 9 The presence of 
this compound probably accounts for the reduction of the silver 
image in a solution of sodium hydrosulfite (experiment 15, Table XI). 

If the addition compounds between sulfur dioxide and hypo are 
oxidizing agents for silver, the reaction represented by the following 
equation II readily explains why the factors previously mentioned 
control the rate of reduction of the silver image in a solution of acid, 
sulfite, and hypo. 

(77.) H,S,0, + H 2 SO, 3= H 2 S203.S0 2 + H 2 O 



Mar., 1932] REDUCING ACTION OF FIXING BATHS 393 

The application of the mass action law to the equation indicates 
that the formation of H^CMSC^) depends upon the acidity, and the 
concentrations of sulfite and hypo. The tendency to form this 
compound increases with an increase in the acidity, and the concentra- 
tion of sulfite and hypo, and hence causes an increase in the rate of 
reduction. A corresponding decrease in the acidity, or the concentra- 
tion of sulfite and hypo, decreases the concentration of the compound 
which also decreases the rate of reduction, which is in accord with 
experimental evidence. 

Bassett and Durrant 9 have shown that methylene blue acts as an 
oxidizing agent in the presence of hypo solutions which explains the 
fact that the dye increases the rate of reduction of the silver image. 
The reaction may be represented by the following equation: 

(7/7.) 2 Ag + C 16 H 18 N 3 SC1 + H 2 S 2 O 3 ;= C 16 Hi 9 N 3 S + Ag 2 S 2 O 3 + HC1 
Methylene Leuco 

Blue Base 

The silver thiosulfate formed in the reaction readily dissolves in 
the excess hypo present. 

Silver halides decrease the rate of reduction of the silver image in 
the fixing baths, probably owing to the formation of complex silver 
anions with the thiosulfate ion, thereby saturating the solution with 
silver. 

The reaction involved in the reduction of the silver image in a solu- 
tion of plain hypo is probably different from that in acid fixing baths, 
since experimental evidence indicates that the rate of reduction for a 
given hypo concentration is only affected by oxidizing agents and 
oxygen. 

Oxygen possibly converts the silver image into silver oxide which is 
readily dissolved by the hypo. 

The reactions may be represented by the following equa- 
tions: 

(77.) 4Ag + 2 ;= 2Ag 2 

(F.) Ag 2 O + Na 2 S 2 O 3 + H 2 O ?=i 2 NaOH + Ag^Os 

The silver thiosulfate formed in equation V readily dissolves in the 
excess hypo present. Equation V also indicates that the solution be- 
comes alkaline when silver oxide is dissolved by hypo which is in ac- 
cord with the experimental facts. 



394 H. D. RUSSELL AND J. I. CRABTREE [j. S. M. p. E. 

VII. SUMMARY 

The object of this investigation was to determine some of the factors 
which control the rate of reduction of the silver image in fixing baths. 

(1) The degree of reduction in a given time was determined for 
images obtained from various emulsions bathed in different fixing 
baths. The emulsions tested included motion picture panchromatic 
negative film type 2, emulsion 1218, and supersensitive panchromatic 
negative film, emulsion 1217, motion picture positive film, emulsion 
1301, motion picture negative film, emulsion 1201, motion picture 
duplicating negative films, emulsions 1505 and 1503, and motion 
picture duplicating positive film emulsion 1355. 

The fixing baths tested were the F-l, the F-2, F-U, F-16, and 7^-23, 
and several experimental formulas. 

(2) The rate of reduction in a given fixing bath was greater with 
images from fine grained emulsions than with coarser grained materials. 
The fixing bath having the lowest rate of reduction was the F-2 
formula, while the highest rates of reduction were obtained with fixing 
baths containing a relatively high concentration of sulfite and acid. 

(3) The rate of reduction increased with an increase in temperature. 

(4) The factor which affected the rate of reduction to the greatest 
degree in an ordinary acid fixing bath was the acidity of the bath. 
For a given bath the rate increases rapidly for pH values below 4.0. 

(5) The rate of reduction was increased for pH values less than 4.0 
with an increase in either the sulfite or hypo concentration. The rate 
of reduction was decreased with concentrations of hypo greater than 
30.0 per cent. 

(6) The exhaustion products which accumulate in a fixing bath 
such as silver halides and developer decreased the rate of reduction. 
Developer oxidation products which also accumulate to a small ex- 
tent did not affect the rate of reduction. 

The rate of reduction in an exhausted F-2 fixing bath from which 
the silver had been removed by an electrolytic process was less than 
in a fresh bath. 

(7) Ammonium chloride, potassium bromide, and potassium iodide 
increased the rate of reduction, while ammonium sulfate, sodium 
chloride, sodium sulfate, glycerin, and sugar produced the opposite 
effect. 

(8) Oxygen and oxidizing agents such as the peroxides have no 
apparent effect on the rate of reduction in highly acid fixing baths. 
The tests indicated, however, that the presence of oxygen increased 



Mar., 1932] REDUCING ACTION OF FIXING BATHS 395 

the rate of reduction in fixing baths containing a low concentration of 
sulfite and acid such as the F-2 formula and was largely responsible 
for the reduction in solutions of plain hypo. 

(9) From a theoretical standpoint most of the factors which control 
the rate of reduction in an acid fixing bath can be accounted for by 
assuming that an oxidizing agent for silver is formed by reaction of 
the hypo and the sulfite. The general formula for such compounds 
is represented by H^OaCSOs)^ and they have been shown 9 to exist in 
an acid solution of sulfite and hypo. 

Other sulfur compounds as well as the halides formed similar addi- 
tion compounds which did not attack the silver image, with the excep- 
tion of the iodide and the hydrosulfite compound. In solutions of these 
compounds, however, the reduction might have been due to the H 2 S2O 3 - 
(80)2) # complex present as an impurity, or as a decomposition product. 

VIII. PRACTICAL RECOMMENDATIONS 

The extent of the reducing effect of fixing baths on the silver image 
during the progress of fixation is greater than has generally been 
supposed. For example, in sensitometric work it is inadvisable to 
prolong the fixation of motion picture positive film in the average 
fresh potassium alum fixing bath beyond 5 minutes at 65F. and 
with certain highly acid chrome alum baths a measurable degree of 
reduction occurs even in this short space of time. 

Since little or no reduction of the image occurs in an alkaline hypo 
solution, sensitometric tests should be checked against images fixed 
in a 25 per cent solution of hypo containing 1 per cent of sodium 
carbonate (anhydrous). The film should be rinsed in water and 
agitated on first immersing in the bath in order to prevent the forma- 
tion of dichroic fog. 10 

In regular laboratory work the degree of reduction which takes 
place in the normal time for fixation is usually of no practical impor- 
tance with the baths in common use. In any given bath the rate of 
reduction increases with the acidity, the temperature of the bath, and 
degree of agitation of the film, so that with certain chrome alum baths 
used under tropical conditions, the decree of reduction is excessive, 
especially with fine grained emulsions. For high temperature 
processing, if a minimum of reduction is required the use of a chrome 
alum hardening stop bath after development, followed by a fixing bath 
consisting of plain hypo containing 1 per cent sodium bisulfite, is 
recommended. 1 



396 H. D. RUSSELL AND J. I. CRABTREE [j. s. M. p. E. 

During use, the reducing action of a fixing bath falls off because it 
becomes more alkaline and accumulates silver thiosulfate which tends 
to retard the reduction. 

In order to insure the minimum degree of reduction, therefore, 
baths having a minimum degree of acidity should be used though such 
baths have a short life and often do not harden satisfactory. It is 
therefore necessary to revive such baths either by adding further 
quantities of acid or hardening solution at intervals during use; 
otherwise, if the film is not rinsed in water before fixing an objection- 
able sludge will form in the fixing bath. 11 

The desirable range of acidity lies between pH values of 4.0 and 4.5. 
At higher values the bath does not harden, and below this there is 
danger of reduction of the image. 

Exposure of the film to air during fixation has little or no effect with 
acid baths, except those containing a relatively low concentration of 
sulfite and acid, in which case the rate of reduction is greatly increased. 
Air also accelerates the rate of reduction in solutions of plain hypo 
which are seldom used in practice. 

The addition of restraining agents such as sodium sulfate, glycerin, 
and sugar to the acid fixing bath decreases the degree of reduction but 
their use is not recommended because they also decrease the rate of 
fixation. 

In some laboratories the acidity of the fixing bath is maintained by 
passing sulfur dioxide gas into the bath. Under these conditions, if 
an excess of the gas is used, a strongly reducing fixing bath is pro- 
duced. 

The nature of the reduction with the negative emulsions tested was 
found to be almost strictly proportional and some of the more active 
baths enumerated could therefore be used advantageously for reduc- 
ing the contrast of photographic images. 

REFERENCES 

1 CRABTREE, J. I., AND RUSSELL, H. D.: "Some Properties of Chrome Alum 
Stop Baths and Fixing Baths," Parts I and II, /. Soc. Mot. Pict. Eng., 14 (May, 
1930), p. 483; (June, 1930), p. 667. 

2 HICKMAN, K. C. D., SANFORD, C., AND WEYERTS, W.: "The Electrolytic 
Regeneration of Fixing Baths," /. Soc. Mot. Pict. Eng., 17 (Oct., 1931), p. 568. 

3 CRABTREE, J. I.: "A Method of Producing Reversed Dye Images," Photo. 
Era, 46 (1921), p. 10. 

4 FOERSTER, F., AND HoRNic, A.: "The Polythionic Acids," Z. anorg. Chem., 
125 (1922), p. 86. 



Mar., 1932] REDUCING ACTION OF FIXING BATHS 397 

5 FOERSTER, F.: "The Formation and Decomposition of Polythionates," 
Z. anorg. Chem., 139 (1924), p. 246; Z. anorg. Chem., 144 (1924), p. 337. 

6 FOERSTER, F., AND MOMMSEN, E. T.: "Thiosulfates," Ber.,57 (B) (1924), 
p. 258. 

7 FOERSTER, F., BROSCHE, A., AND NORBERG-SCHULTZ, C.: "Sodium and 
Potassium Salts of Sulfurous Acid," Z. physik. Chem., 110 (1924), p. 435. 

8 FOERSTER, F., LANGE, F., DROSSBACH, O., AND SEIDEL, W.: "The 
Decomposition of Sulfurous Acid and Its Salts in Aqueous Solutions," Z. anorg. 
Chem., 128 (1923), p. 245; FOERSTER, F., AND KUBEL, K.: "The Decomposition 
of Sulfites at Red Heat," Z. anorg. Chem., 139 (1924), p. 261; FOERSTER, F., 
AND VOGEL, R.: "The Behavior of Sulfurous Acid toward Thiosulfuric Acid," 
Z. anorg. Chem., 155 (1926), p. 161; FOERSTER, F., AND CENTNER, R.: "The 
Action of Sulfites on Polythionates," Z. anorg. Chem., 157 (1926), p. 45; FOERS- 
TER, F., AND HAMPRECHT, G.: Z. anorg. Chem., 158 (1926), p. 277; FOERSTER, 
F., AND HAUFE, E.: "The Auto- Decomposition of Aqueous Hydrogen Sulfite 
Solutions," Z. anorg. Chem. 177 (1928-29), p. 17; FOERSTER, F., AND KIRCHEISEN, 
E.: "The Interaction of Hydrogen Sulfite and Hydrosulfite," Z. anorg. Chem., 
177 (1928), p. 42; FOERSTER, F.: "The Inter-Relationship of the Sulfur Acids," 
Z. anorg. Chem., 177 (1928-29), p. 61. 

9 BASSETT, H., AND DURRANT, R. G.: "The Inter-Relationships of Sulfur 
Acids," /. Chem. Soc. (1927), p. 1401. 

10 CRABTREE, J. I.: "Stains on Negatives and Prints," Amer. Ann. Phot., 35 
(1921), p. 204. 

11 CRABTREE, J. I., AND HARTT, H. A.: "Some Properties of Fixing Baths," 
Trans. Soc. Mot. Pict Eng., 13 (1929), No. 38, p. 364. 



ABSTRACTS 

The dews of the readers of the JOURNAL relative to the usefulness to them of the 
abstracts regularly published in the JOURNAL will be appreciated. Favorable views 
are of particular interest. In the absence of a substantial body of opinion to the 
effect that these abstracts are desired by the membership, their discontinuance may be 
considered. 

Experiments with Visual Aids in High School Classes. W. LEWIN. Visual 
Instr. News, 5, Nov., 1931, p. 9. Another quite independent experiment to test 
the efficacy of motion pictures in teaching, the subject being high school physics. 
Preliminary intelligence, reading, and physics tests showed the control groups to 
have a very slight advantage. Motion pictures were presented to the experi- 
mental group during their preparatory period while the control group met for 
supervised study. It was concluded that motion pictures impart more informa- 
tion in a given time and also contribute to retention of information. The gain in 
the test grades of the experimental group over the control group was three times 
the standard error while at the end of the term, 50 per cent more pupils of the ex- 
perimental group passed the course. R. P. L. 

A Modern Theater for the Classics. N. BEL GEDDES. Theater Management, 
26, Nov., 1931, p. 8. A theater specially designed for the staging of Dante's 
Divine Comedy at the Chicago World's Fair has a seating capacity of 5000 and is 
similar to the ancient Greek theater. Its plan is a half -circle facing the stage 
without balconies or galleries. No proscenium or curtain divides the auditorium 
from the stage. The absence of balconies and galleries allows a steeper ramp and 
better vision from all seats. The stage is circular and composed of steps. In the 
center is a pit, at the far side of which the slope rises to a height of 50 feet. On the 
near side, the slope terminates in a ledge only one-fourth as high, which steps 
down toward the audience in a series of terraces until it reaches the level of the 
bottom of the pit where it terminates in a valley running half-way around the 
circle. A 7-foot wall separates the valley from the audience. Mention is made 
of two other theaters also planned for the World's Fair in which the absence of 
transverse aisles is notable, the rows of seats being given liberal spacing instead. 

L. E. M. 

Room Noise Reduction for Improved Sound Reception. V. A. SCHLENKER. 
Theater Management, 26, Nov., 1931, p. 3. A study of the relations of speech, 
music, and room noise in the theater indicates that the noise level should be re- 
duced below 30 decibels for the speech, and music must be uncomfortably loud to 
be heard above the noise level of 40 to 50 decibels. Excessive treatment of the 
theater proper should be avoided in view of a possible interference with the proper 
reverberation period which is considered essential to the proper diffusion of sound 
to all parts. The room noise can generally be controlled to suitable value by de- 
creasing street and lobby noise through maximum treatment in the lobby and 
foyer. L. E. M. 

398 



ABSTRACTS 399 

A Clockwork Driven Slow-Motion Camera. Kinemat. Weekly, 178, Dec. 17, 
1931, p. 38. A new type of slow-motion picture camera which is actuated by 
clockwork is claimed to expose 100 feet of 35-mm. film with one winding of the 
mechanism. The speed can be varied from 40 to 120 frames per second and a 
reversed fitting allows dissolving to be carried out while the film is being exposed. 
A pick-up speed has been developed which permits only 18 inches of film passage 
before full rate is obtained. Stopping and starting can be accomplished with a 
loss of less than 2 feet of film. -A reflex focusing device permits accurate focusing 
when taking close-ups, and the enclosed view finder is fitted with a device to allow 
for parallax when the object is near the camera. 

A standard speed camera designed similarly to the slow-motion model, but 
capable of exposing 200 feet of film at speeds from 10 to 24 frames, has also been 
introduced. A special tripod is used with these models. C. H. S. 

Effect Lighting. J. H. KURLANDER. Theater Management, 27, Jan., 1932, p. 
10. Suitable lighting effects are proposed for the theaters having a straight sound 
picture program so as to relieve the show of monotony. A description of equip- 
ment required for effect lighting is given. The uses of effect projectors, shutters, 
framing devices, masks, slides, special screens, etc., for producing different effects 
are discussed. Color effects, animated scenic effects, silhouettes, trick effects, 
and others may be used as the occasion suggests. W. J. W. 

Diminishing the Fire Hazard. J. J. GREILSHEIMER. Theater Management, 27, 
Jan., 1932, p. 16. The use of concrete vaults or sheet metal lockers, even though 
equipped with sprinkler systems and vents, is deemed inefficient in preventing 
film fires because of the large quantity of film concentrated in one compartment. 
Several requirements for a safe and efficient film storage cabinet are enumerated. 
A description is given of a cabinet designed to meet these rigid requirements. The 
cabinet is constructed in sections featuring individually insulated and ventilated 
compartments of 10 pounds capacity which are sealed tightly with automatically 
closing and latching doors. A number of fire tests were carried out on the cabinet 
filled with film to determine its safety. Detailed results of the tests are given. 

W. J. W. 

Advances in Sound Reproduction Demonstrated to Motion Picture Engineers. 
Theater Management, 27, Jan., 1932, p. 5. Reproductions of organ, orchestral, 
and vocal music, which closely approached the quality and volume of the original, 
were effected by the use of disk records cut by the vertical method. This method 
employs grooves which vary in depth instead of wavering back and forth along the 
spiral path as in the commonly used lateral method. The moving element of the 
electrical reproducer is made of light-weight materials so that it is able to follow 
vibrations up to 10,000 per second with fidelity. A tiny permanent sapphire 
point is used which rides smoothly up and down in the grooves. Finished 
records are pressed in cellulose acetate which has a surface of extremely fine tex- 
ture. Mr. H. A. Frederick of the Bell Telephone Laboratories made the demon- 
stration. W. J. W. 

Television Talkiola. Theater Management 26, Nov., 1931, p. 34. This ap- 
paratus incorporates mechanisms for producing six different types of entertain- 
ment within a single cabinet, namely, television with synchronized sound, talking 
motion pictures (16-mm. or silent pictures), phonograph, short wave radio, and 



400 ABSTRACTS [j. S. M. p. E. 

standard broadcast radio. A Vis-horsepower synchronous motor operates the 
perforated scanning disk used for television, giving a 6- by 8-inch picture. Rear 
projection is used for the 16 mm.-projector. G. E. M. 

New Type Record. Theater Management, 26, Nov., 1931, p. 34. This new disk 
record is made of much thinner material and is much less easily broken than the 
old type shellac record. Although only 12 inches in diameter, as compared with 
the older 16-inch record, the new disk will record sufficient sound for 1000 feet of 
film. This has been accomplished by employing a lower amplitude of recording, 
smaller grooves, and by placing the grooves nearer together. G. E. M. 

Novel Loud Speaker. R. H. CRICKS. Kinemat. Weekly, 173, July 9, 1931, p. 
69. New principles are claimed in the construction of a novel loud speaker which 
has recently been demonstrated in London. Known as the Cinemavox, it is 
stated to combine the principles of the piano and violin by providing a large 
tuned area for the dissemination of sound. A number of speaker armatures are 
distributed at the back of a sounding board some 5 feet square, and are connected 
to struts, which are parts of various wooden sections, each having its own natural 
resonance frequency. A frequency range of from 13 cycles to 17,000 cycles with 
extremely even response is claimed. The sound output is stated to be almost 
non-directional. Kodak Abstract Bulletin 

New "Jofa" Studio. P. HATSCHEK. FUmtechnik, 7, Sept. 19, 1931, p. 6. A 
description is given of the new "Jofa" sound film studio of Jahannisthal, Berlin, 
which is the most up-to-date in the city. There are three large studios, 840, 
1155, and 840 square meters in area, each associated with a smaller studio, res- 
pectively, 480, 480, and 450 square meters in size, and a large number of dressing- 
rooms, and smaller rooms for operators, technicians, actors, etc. There are two 
studios for re-recording, dubbing, and synchronizing, four projection rooms, two 
cutting rooms, and a number of work-shops. Thirty thousand square meters of 
land are available for outdoor work, and an additional seventy thousand meters 
(the local aerodrome) are at hand if required. The three large studios have 
enormous sliding doors opening on the outside lots. This provides a natural 
background for studio sets, if desired, and permits a continuation of the studio 
action outdoors. For sound-proofing, air spaces are provided between studios, 
the floors are insulated from the walls by coke-ash, walls and doors are all double, 
and are packed with sound-absorbing material. Doors are provided with a 
novel "double-fold system" which is described, and there is a new treatment of 
the roof. The electrical supply and the projection and cutting rooms are also 
described. A pool, 35 by 15 meters wide and 2.5 meters deep, is provided. 

Kodak Abstract Bulletin 

Modern Effect Lighting. J. H. KURLANDER. Mot. Pict. Proj., 5, Jan., 
1932, p. 18. A descriptive article on the production of stage and screen light- 
ing effects, including information on lamp and lens equipment, types of screen 
and screen materials, and the use of color filters, slides, and design glasses. 

A. A. C. 

Projected Background Cinematography. R. G. FEAR. Amer. Cinemat., 12, 
Jan., 1932, p. 11. A method of composite photography is described in which 
the foreground action takes place in front of a screen placed so as to receive from 
a projector an image of the background desired. Translucent screens in back of 



Mar., 1932] ABSTRACTS 401 

the action are now often used for this purpose with a standard camera and pro- 
jector. The background picture must be absolutely steady on the screen, 
illuminated to the highest possible extent, and must be synchronized with a 
camera shutter if good results are to be secured. After a discussion of means of 
fulfilling these requirements, the author suggests modifications that may prove 
useful, and gives a list of patents relating to the process. A. A. C. 

New Filters for Exterior Photography with Super- Sensitive Film. EMERY 
HUSE AND GORDON A. CHAMBERS. Amer. CinemaL, 12, Dec. 1931, p. 13. Two 
new filters, the 3 N5 and 5 N5, are combinations of yellow dyes with a neutral 
density filter of 32 per cent transmission. They combine, in a single unit, a 
means of decreasing exposure and a color filter suited to the super-sensitive emul- 
sion. This means of reducing light intensity has been found preferable to using 
a lens diaphragm or a change in shutter opening A. A. C. 

Projector Drive Motors. ALBERT PREISMAN. Mot. Pict. Proj., 5, Jan., 
1932, p. 10. Since the advent of sound, the projector drive motor has assumed a 
greater importance than ever before. Ease and precision of control, affording a 
constant and definite speed, are imperative. The article discusses the underlying 
principles of the common types of projector motors and explains how the new 
demands are met in modern motor design. A. A. C. 

Reverberation Time Measurements in Coupled Rooms. CARL F. EYRING. 
/. Acoust. Soc. Amer., Ill, No. 2, Part I, Oct., 1931, p. 181. The paper pre- 
sents experimental data on the decay of sound intensity level in acoustically 
coupled rooms, together with a theoretical study of the subject. 

The type of problem investigated is illustrated by one of the experiments, 
which was a study of the sound decay in an enclosure which consisted of a small 
live room connecting through an open window into a large dead room. Data 
were taken with the sound source in the large room and microphone in the small 
room, and vice versa, and with both source and micrpohone in each room. Com- 
binations of other types of rooms are included. 

Theoretical equations of decay for acoustically coupled rooms are developed, 
and are applied to describe the data. The application of these equations to an 
idealized theater is shown. W. A. M. 

Audible Frequency Ranges of Music, Speech, and Noise. W. B. SNOW. /. 
Acoust. Soc. Amer., Ill, No. 1, Part 1, July, 1931, p. 155. "The program of 
listening tests described in this paper was undertaken primarily to establish the 
audible frequency ranges of the sounds most often encountered in sound repro- 
duction. ..." The sound sources studied included twenty separate musical in- 
struments, an orchestra, male and female speech, and certain noises. 

Qualitative observations by the crew of listeners are tabulated for each sound 
source. Quantitative results are given in a table. Two general conclusions are 
as follows: "An upper cut-off of 10,000 cycles did not affect the tone of most of 
the instruments to a marked extent, but every instrument except the bass drum 
and tympani was affected by the 5000 cycle cut-off. A frequency range of 100 to 
10,000 cycles was shown to be entirely satisfactory for speech." ". . . . trans- 
mission of the entire audible range would seem much more important for noise 
reproduction than for reproduction of musical sounds." 

The paper contains a great amount of experimental data. W. A. M. 



402 ABSTRACTS 

Plane Sound Waves of Finite Amplitude. R. D. FAY. /. Acoust. Soc. Amer., 
Ill, No. 2, Part I, Oct., 1931, p. 222. The principal object of the analysis is 
to find the change in type of periodic plane waves of sound of finite amplitude 
propagated in free air. 

A solution of the exact equation of motion is obtained as a Fourier series. Due 
to the non-linear relation between pressure and specific volume there is found to 
be a gradual transfer of energy from components of lower frequency to those of 
higher frequency. Since the effect of viscosity is to attenuate the higher fre- 
quency components more than the lower, there is always a wave form having the 
harmonic components in a stable relation such that the decrease in relative mag- 
nitude of any component due to viscosity is compensated by the relative increase 
due to non-linearity. The conditions for stability vary with intensity. There is 
therefore no permanent wave form, but the stable wave will change its form more 
gradually than any other wave of the same intensity and wavelength. The 
change in type of any wave is toward this stable form. There is a marked de- 
parture from the sinusoidal in the stable type even for waves of very moderate 
amplitude. AUTHOR 

A Planetary Reduction Gear System for Recording Turntables. A. V. BED- 
FORD. /. Acoust. Soc. Amer., Ill, No. 2, Part I, Oct., 1931, p. 207. "The 
present paper has two objects: to present an example justifying the use of a de- 
tailed numerical application of electrical circuit analysis to mechanical rotational 
systems, and to describe a new planetary turntable drive system that promises 
increased steadiness." 

The conclusion of an analysis of a simple gear system is that, "... the error of 
the turntable position at any moment is about as great as the fundamental error 
in the angular tooth pitch in the lowest speed gear." 

In the planetary gear system described no gear runs as slow as 33 Vs rpm. with 
respect to its meshed mate, and also no gear in the system runs at a speed lower 
than 375 rpm. Therefore, disturbances due to errors in gears and irregularities in 
bearing friction are of a relatively higher frequency than in a simple gear system 
and consequently can be more easily filtered out. 

An experimental model of a planetary gear system drive "exhibited less than 
0.03 per cent variation in turntable speed at turntable revolution frequency." 

W. A. M. 



BOARD OF ABSTRACTORS 

BROWNELL, C. E. MACFARLANE, J. W. 

CARRIGAN, J. B. MACNAIR, W. A. 

COOK, A. A. MATTHEWS, G. E. 

CRABTREE, J. I. McNicoL, D. 

HAAK, A. H. MEULENDYKE, C. E. 

HARDY, A. C. MUEHLER, L. E. 

HERRIOT, W. PARKER, H. 

IRBY, F. S. SANDVICK, O. 

IVES, C. E. SCHWINGEL, C. H. 

LOVELAND, R. P. SEYMOUR, M. W. 

WEYERTS, W. 



ABSTRACTS OF RECENT U. S. PATENTS 

The views of the readers of the JOURNAL relative to the usefulness to them of the 
Patent Abstracts regularly published in the JOURNAL will be appreciated. Favorable 
views are of particular interest . In the absence of a substantial body of opinion to 
the effect that these Patent Abstracts are desired by the membership, their early dis- 
continuance may be considered. If, after two weeks from the date of mailing the 
March issue of the JOURNAL, no letters concerning the continuance of the depart- 
ment will have been received, the Patent Abstracts will be discontinued. 

1,828,798. Film Treating Apparatus. G. C. BEIDLER. Oct. 27, 1931. The 
film is delivered edgewise to means for removing the film from the receptacle in 
which a submerging device is located and other guiding means operate to prevent 
lateral movement of the film as it is being moved. Means are provided for regu- 
lating tension or pressure on the film by coacting rollers which operate to move 
the film and at the same time exert pressure upon the film to remove fluid, in 
order to prevent film from carrying an excess amount of fluid from the receptacle 
in which the film was treated. At the bottom of the coils where they coact, 
means are provided for moving the film to eject it from a receptacle, an assembly 
of rollers and conveying bands being provided for continuously directing the film. 

1,828,749. Motion Picture Screen. A. L. RAVEN. Oct. 27, 1931. The pro- 
jection screen comprises a plurality of wavy horizontal strips arranged in overlap- 
ping relation with the hollows of the waves of adjacent strips opposite one another 
and forming sound passages extending upwardly from the rear toward the front of 
the screen between the strips. The sound from the sound reproducer behind the 
screen freely passes through the screen at the same time that a proper reflection 
surface is provided for the screen. 

1,828,768. Film Guide. A. DINA. Assigned to International Projector 
Corp. Oct. 27, 1931. One set of guide members is rigidly mounted for positively 
locating the film edge with respect to the projection aperture and comprises a 
plurality of sections spaced longitudinally of the film for permitting dust and 
accumulations of foreign material to escape therebetween. The other set of 
guide members comprises a plurality of disks rotatably mounted with their axes 
transverse to the film and held in firm engagement therewith by means of suitable 
spring members. The disks are capable of rotating as the film is moved through 
the projection head thereby eliminating sliding friction and reducing the wear on 
the film. 

1,828,867. Scanning Device. C. FRANCIS JENKINS. Assigned to Jenkins 
Laboratories. Oct. 27, 1931. The film image is enlarged by projection and 
directed through a scanning disk thereby permitting (1) the apertures in the 
scanning disk to be larger, so that diffraction bears a lesser relation to the aper- 
ture area; (2) the disk may be positioned in a free air, removed from the proximity 
of the film, and, therefore, does not clog up with dirt and/or oil; and (3) the 
apertures may be made square, increasing the light efficiency. 

1,828,875. Electrooptical Translation System. C. H. W. NASON. Assigned 

403 



404 PATENT ABSTRACTS [J. S. M. p. E. 

to Jenkins Television Corp. Oct. 27, 1931. A method of employing photo- 
electric variations to control the resonance characteristic of an oscillatory circuit 
supplied from a source of carrier current. The frequency spectrum of the trans- 
mitted carrier waves is substantially independent of the frequency variations of 
the light impulses incident upon the light-sensitive device under control of a film. 
The light passing through each elemental area of the film is projected upon the 
photoelectric cell, preferably of the Elster-Geitel type, comprising a light-sensitive 
electrode and another electrode. The electrostatic capacity of such a cell under- 
goes variations in value as the coating is subjected to different degrees of illumi- 
nation. 

1.828.940. System for Correcting Sound Records. R. J. POMEROY. Oct. 27, 
1931. A method and system for making a distortion corrected record, by intro- 
ducing to the original record correction distortions that are compensatory of, or 
have a neutralizing effect on, the distortions which are introduced by reproduc- 
tion. This is done by recording the distorted reproduced sound, and utilizing 
this distorted record to modify the original record in such a manner that the dis- 
tortive effects of the system are compensated in the modified record, and accurate 
reproduction is thus obtainable. 

1.828.941. System for Correcting Sound Records. R. J. POMEROY. Oct. 27, 
1931. A method and system for making a distortion corrected record, and this is 
done by introducing, to the record, correction distortions that are compensatory 
of, or have a neutralizing effect on, the distortions which are introduced by repro- 
duction. A sound current representing the distortion record is combined with a 
sound current representing the original undistorted record, and this combination 
is so effected that the resultant current carries variations which represent only the 
difference between the two records or, in other words, the distortion. A record 
of this current may be made upon a film and subsequently printed above an ori- 
ginal record. In either case, the result is a distortion corrected record from which 
sound may be finally reproduced without the distortions of recording and repro- 
duction. 

1,828, 942. Production of Corrected Sound Records. R. J. POMEROY. Oct. 
27, 1931. A method and system for making a distortion corrected record, and 
this is done by introducing to the record correction distortions that are compensa- 
tory of, or have a neutralizing effect on, the distortions which are introduced in 
recording and reproduction. This is accomplished in the present instance by 
making a photographic distortion corrected sound record, or photographic sound 
record compensated for distortions, and from this making a distortion corrected 
mechanical record from which distortionless reproduction is obtainable. 

1 ,828,974. Photographic Film with Visible Reproducible Inscriptions. H. LUM- 
MERZHEIM AND E. ScHNiTZLER. Assigned to Agfa Ansco Corp. Oct. 27, 1931. 
An ink is provided for continuous printing on a photographic film, the ink com- 
prising a mixture of cerasine-red in glycol acetate. A photographic film provided 
on the rear side with inscriptions by means of the said dye-ink may be polished as 
usual in the photographic film industry; it may be exposed, developed, and fin- 
ished in the usual manner without fading of the impressed symbols. 

1,829,095. Film Reel. W. G. KING AND M. E. KRAUSE. Oct. 27, 1931. An 
endless film may move continuously or intermittently in a continuous path from 
the inner convolution of a roll of film revolving about the circularly grouped rollers 



Mar., 1932] PATENT ABSTRACTS 405 

onward through the mechanism of a projector and past the lens and back onto 
the outer convolution of a roll of film without undue strain or intricate twists or 
loops in the film by the provision of yieldable film guides. 

1,829,103. Loading Device for View Taking Cinematographic Apparatus. 
A. N. MERLE. Assigned to Pathe Cinema, Anciens Etablissements Pathe Frdres. 
Oct. 27, 1931. The cover is provided with a bevelled part corresponding to that 
of the cover and is formed on the face of the loading case coacting with the cover. 
This bevelled part is situated outwardly of the film-holding chamber. The said 
bevelled part may extend upon the whole periphery of the said chamber or upon 
only a certain portion thereof. The loading case may be readily opened to allow 
access to the hollow interior of the box for the insertion or removal of the film. 

1,829,121. Sound Recording Apparatus. E. R. VINSON. Oct. 27, 1931. An 
electromagnetic vibratile device for moving a light valve in the form of a V- 
shaped notch in the path of a beam of light for varying the exposure of the film 
according to impressed sound vibrations. 

1,829,359. Picture Projecting Machine Cabinet. R. W. KITTREDGE. Oct. 
27, 1931. Cabinet for a motion picture projecting machine, and a projection 
screen and stand therefor which is removably stored on the cabinet in such a 
manner as not to decrease materially the space afforded in the cabinet for the re- 
ception or storage of other articles such as a projecting machine, related appara- 
tus, and film, which permits the quick and convenient storing of the screen and 
stand on the cabinet and removal of the same therefrom, and which does not de- 
tract from the appearance of the cabinet or require an unattractive shape thereof 
for use in the home to form an attractive and convenient article of furniture. 

1,829,475. Projection Lamp Holder. G. H. CUSHING. Oct. 27, 1931. A 
tubular holder for an incandescent lamp by means of which a standard electric 
light bulb may be positioned in an accurately designed reflector, so that the fila- 
ment of the bulb will be located at the focus of the reflector. 

1,829,482. Motion Picture Film Reel. Oct. 27, 1931. A. C. Hayden. A 
film reel for motion picture apparatus comprising a pair of plates and a hub be- 
tween said plates adapted to have a film wound thereon, one of said plates having 
a hole therein and the other of said plates having a cup integral therewith, pressed 
thereform and in axial alignment with the hole, said hole and cup being adapted 
to receive a spindle of motion picture apparatus, said hole being formed for driving 
engagement with the spindle, and said cup insuring application of the reel to the 
spindle with the hole plate in advance of the cup plate. 

1,829,633. Taking or Projecting Panoramic Views or Views Extending in 
Height. H. CHRETIEN. Assigned to Societe Anonyme Francaise Dite Societe 
Technique D'Optique et de Photographic. Oct. 27, 1931. Method of photo- 
graphing or projecting which consists of reducing optically the space occupied by 
the images on a sensitized surface, by compressing them in one single direction, 
either in height, or in width, or in any inclined direction selected, this result being 
obtained by disposing, in front of the photographing objective, a special optical 
combination, referred to as a local anamorphoser, suitably oriented about the 
optical axis of the objective. The process also consists in restoring or projecting 
these images through an optical combination similar to that which has served for 
obtaining them and similarly directed, which has the result of reestablishing the 



406 PATENT ABSTRACTS 

images in their exact proportions on a screen of suitable dimensions and arrange- 
ment. 

1,829, 634. Optical Compression of Film Pictures. H. CHRETIEN. Oct. 27, 
1931. A film which includes a series of pictures thereon of uniform dimensions 
and proportions which are optically compressed, some in one dimension and some 
in another, so as to obtain when projected and restored views which are consider- 
ably extended but only in the one dimension or the other. 

1,829,791. Device for Recording Sound on Film. H. A. DEVRY. Assigned to 
Q. R. S.-De Vry Corp. Nov. 3, 1931. Incandescent lamp having a bulb, part of 
which is opaque except for a minute slit in the tip end thereof. The lamp is 
adapted to be positioned with respect to feeding or winding mechanism for the 
film so that the beam of light emanating from the slit strikes against one of the 
side margins of the film and forms on the film, as the latter is driven by the feed 
mechanism, an exposed portion of strip-like conformation. 

1,829,912. Sound Picture Film and Method of Making the Same. D. G. 
SHEARER. Assigned to Metro-Goldwyn-Mayer Corp. Nov. 3, 1931. Con- 
tinuous picture film and sound record comprising a strip of film bearing pictures 
between rows of sprocket holes made therein, and a continuous photographic 
sound record on film stock attached to one longitudinal edge of the picture film, 
one edge of the picture film being stepped and the stepped edges cemented to- 
gether whereby the combined picture film and sound record are of substantially 
equal thickness transversely thereof. 

1,830,082. Color Attachment for Cinema Projectors. W. R. BECKLEY, A. E. 
CHURCH AND J. F. MERKEL. Assigned to Beckley and Church, Inc. Nov. 2, 
1931. A rotatable disk is placed upon the front of the projector and arranged in 
front of the lens and adapted to present various differently colored transparent 
segments thereof in the axis of the lens, selectively, so that the projected rays will 
be colored or filtered in a manner such as will protect the eye of the observer from 
the glare of the image as projected upon the screen and also when desired to im- 
part a colorful effect simulating, for instance, moonlight, twilight, etc. 

1,830,121. Color Attachment for Cinema Projectors. J. F. MERKEL. As- 
signed to Beckley & Church, Inc. Nov. 3, 1931. A carrier for a lens plate is 
mounted on the projector and a vari-colored ray screen rotatably and reversibly 
mounted on the shaft in operative relation to the lens for the reproduction of 
images in color. 

1,830,158. Film Trap and Film Trap Door. A. DINA. Assigned to The Pre- 
cision Machine Co., Inc. Nov. 3, 1931. Construction of film guide and film 
trap door in which the door is closed upon the film against impact absorbing 
means which prevents transmission of shocks to parts of the projector. A resilient 
contacting pad is provided against which the door is moved to closed position. 

(Abstracts compiled by John B. Brady, Patent Attorney, Washington, D. C.) 



SOCIETY OF MOTION PICTURE 
ENGINEERS 

OFFICERS 
1931-1932 

President 
A. N. GOLDSMITH, Radio Corporation of America, New York, N. Y. 

Past-President 
J. I. CRABTREE, Eastman Kodak Company, Rochester. N. Y. 

Vice-Presidents 

W. C. HUBBARD, General Electric Vapor Lamp Co., Hoboken, N. J. 
E. I. SPONABLE, Fox Film Corp., New York. N. Y. 

Secretary 
J. H. KURLANDER, Westinghouse Lamp Co., Bloomfield, N. J. 

Treasurer 
H. T. COWLING, Eastman Teaching Films, Inc., Rochester, N. Y. 

Board of Governors 

F. C. BADGLEY, Canadian Government Motion Picture Bureau, Ottawa, Canada 
H. T. COWLING, Eastman Teaching Films, Inc., 343 State St., Rochester, N. Y. 
J. I. CRABTREE, Research Laboratories, Eastman Kodak Co., Rochester, N. Y. 
P. H. EVANS, Warner Bros. Pictures, Inc., 1277 E. 14th St., Brooklyn, N. Y. 
O. M. GLUNT, Bell Telephone Laboratories, New York, N. Y. 
A. N. GOLDSMITH, Radio Corporation of America, 570 Lexington Ave., New 

York, N. Y. 

W. C. HUBBARD, General Electric Vapor Lamp Co., Hoboken, N. J. 
R. F. MITCHELL, Bell & Howell Co., 1801 Larchmont Ave., Chicago, 111. 
J. H. KURLANDER, Westinghouse Lamp Co. Bloomfield, N. J. 
W. C. KUNZMANN, National Carbon Co., Cleveland, Ohio 

D. MACKENZIE, Electrical Research Products, Inc., 7046 Hollywood Blvd., 

Los Angeles, Calif. 
L. C. PORTER, General Electric Co., Nela Park, Cleveland, Ohio 

E. I. SPONABLE, 277 Park Ave., New York, N. Y. 

407 



SOCIETY ANNOUNCEMENTS 

SPRING, 1932, MEETING 

The Spring, 1932, Convention of the Society is to be held at Wash- 
ington, D. C., with headquarters at the Wardman Park Hotel. 
Excellent service is assured and plenty of space is available for 
accommodating the members without crowding. The Congressional 
Country Club and the Indian Spring Country Club are both available 
to the visiting members. In addition, the four tennis courts main- 
tained by the hotel and riding facilities provide additional recreation. 

As the Convention is to be held at the height of the activities of 
the Washington Bi-Centennial celebration, there will be much to 
attract members to Washington in addition to the technical and 
social activities of the Society. Sight-seeing tours will be provided 
for visiting points of historic and diplomatic interest, such as the 
Capitol, the Treasury, the Smithsonian Institution, the Congres- 
sional Library, the Pan-American Building, the new Museum, 
Scottish Rite Temple, Tomb of the Unknown Soldier, the Amphi- 
theater, and other points of interest at Arlington, Mount Vernon, and 
Annapolis. 

An especially attractive program of technical papers is being 
prepared by the Papers Committee, under the chairmanship of 
Mr. O. M. Glunt; and Mr. W. C. Kunzmann and his Convention 
Arrangements Committee are sparing no efforts to make the social 
aspects of the Convention a success. The technical sessions will 
be held in the Little Theater of the Wardman Park Hotel, and special 
film programs for the evenings are being arranged by Mr. J. I. 
Crabtree. 

The semi-annual banquet of the Society will be held in the Gold 
Room of the hotel on Wednesday, May llth, at 7:30 P.M. In 
addition to an attractive and entertaining program, an unusually 
interesting group of speakers is expected to address the members. 

NEW YORK SECTION 

On February llth, the members of the New York Section were in- 
vited by the Illuminating Engineering Society to attend its Febru- 
408 



SOCIETY ANNOUNCEMENTS 409 

ary meeting held at the plant of the Sperry Gyroscope Company, 
Brooklyn, N. Y. A paper entitled "The Theory of the Arc, and the 
Carbon Arc as a Projection Source" was presented by Mr. Bassett, 
of that company. 

The members of the Section were also invited by the New York 
Section of the American Institute of Electrical Engineers to attend 
its meeting held on February 26th, at the Engineering Societies 
Building, New York, N. Y. The meeting was devoted to "The 
New Music of Electrical Oscillations," and included demonstrations 
of the electronic organ-piano, developed by Mr. Benjamin Miessner 
and the Ranger tone electric organ, developed by Captain Richard 
Ranger. Professor Leon Theremin demonstrated the three types 
of theremin the space theremin, the new keyboard theremin, and 
the new fingerboard theremin. 

CHICAGO SECTION 

At a meeting of the Section, held on January 7th, papers were 
presented by Mr. R. W. Fenimore, entitled "Educational and Com- 
mercial Films with Sound on Disk," and by Mr. L. D. Minkler, on 
"Disk Recording for Motion Pictures." 

At another meeting held on February llth, Mr. R. F. Mitchell 
presented a paper entitled "New Improvements in Camera Con- 
struction." 

The next meeting of the Section will be held March 3rd at the 
Electric Association, Chicago, 111. Mr. H. Shotwell will present a 
paper on the subject of "Portable A-C. Amplifiers." 

STANDARDS COMMITTEE 

At two meetings of the subcommittee of the Committee on 
Standards and Nomenclature, which deals with the establishment 
of dimensional standards for 16-millimeter sound film, on January 
28th and February 8th, two lay-outs were made, which are to be 
submitted to the entire Standards Committee for consideration 
and appropriate action. The one lay-out, providing for a single 
row of perforations, is to be submitted for adoption as a recommended 
standard of film lay-out; the other, providing for two rows of per- 
forations, is also to be submitted to the Standards Committee, with 
the suggestion that this be published (somewhat as in the nature of 
a minority report) as a non-recommended standard, to be followed 
if future developments of the art so indicate. 



410 SOCIETY ANNOUNCEMENTS [J. S. M. P. E. 

Drawings of the two lay-outs are being prepared, showing all 
details and tolerances, which will be submitted to the Standards 
Committee at its next meeting, to be held in the near future. Upon 
ratification of these lay-outs, as submitted or modified, they will be 
published in the next succeeding issue of the JOURNAL. 

SOUND COMMITTEE 

At a meeting held on December 10, 1931, an outline of the work to 
be prosecuted by the Committee during the current year was formu- 
lated, and included a considerable amount of study of the acoustical 
properties of auditoriums and studios, with particular reference to 
the influence these properties exert in the recording and reproducing 
of sound. An attempt will be made to define an optimum theater, 
that is, one whose properties may be regarded as reference standards 
which will indicate the factors to be considered in making audi- 
toriums acceptable for the reproduction of sound. Among the other 
items included in the agenda are: (1) the accuracy and application 
of testing methods and formulas; (2) absorption data of acoustic 
materials; (3) wide-range recording and reproducing of sound; 

(4) sources of ambient or interfering noises, and their correction; 

(5) the relation between the acoustical properties of studios and 
theaters; (6) the influence of the light slit and of the methods of 
processing film on the frequency characteristic of reproduction; 
(7) the desirability of increasing the range of volume of reproduction ; 
and (8) variations in negative exposures. 

Various subcommittees have been appointed to study these 
several subjects outlined, the reports of which subcommittees are 
to be submitted at a meeting of the entire Committee in the near 
future. 

JOURNAL AND PROGRESS AWARDS 

At a meeting of the Board of Governors held May 24, 1931, 
it was decided that the following actions of the Board, relating to 
the Journal Award and the Progress Medal, should be published 
annually in the JOURNAL. 

JOURNAL AWARD 

The motion was made and passed that "an award of $100.00 shall 
be made annually, at the Fall Convention of the Society, for the 



Mar., 1932] SOCIETY ANNOUNCEMENTS 411 

most outstanding paper published in the JOURNAL of the Society 
during the preceding calendar year. An appropriate certificate 
shall accompany the presentation. 

"The Journal Award Committee shall consist of not less than six 
Active members of the Society, to be appointed by the President sub- 
ject to ratification by the Board of Governors. The Chairman of the 
Committee shall be named by the President and a two-thirds vote is 
necessary for election to the award. (Proxies are permitted.) 

"The Committee shall be required to make its report to the Board 
of Governors at least one month prior to the Fall Meeting of the 
Society, and the award must be ratified by the Board. A list of 
five papers shall also be recommended for honorable mention by 
the Committee. These rules, together with the titles and authors' 
names, shall be published annually in the JOURNAL of the Society." 

PROGRESS MEDAL 

"The Board of Governors may consider annually the award of a 
Progress Medal in recognition of any invention, research, or develop- 
ment, which in the opinion of the Progress Award Committee shall 
have resulted in a significant advance in the development of motion 
picture technology. 

"The Committee shall consist of not less than six Active members 
of the Society, to be appointed by the President subject to ratifica- 
tion by the Board of Governors. Names of persons deemed worthy 
of the award may be proposed and seconded, in writing, by any 
two Active members of the Society and shall be considered by the 
Committee during the month of June; a written statement of ac- 
complishments shall accompany each proposal. 

"Notice of the meeting of the Progress Award Committee must 
appear in the March and April issues of the JOURNAL. All names 
shall reach the Chairman not later than April 20th. 

"A two-thirds vote of the entire Committee shall be required to 
constitute an award of the Progress Medal. Absent members may 
vote in writing. The report of the Committee shall be presented 
to the Board of Governors for ratification at least one month before 
the Fall Meeting of the Society. 

"Recipients of the Progress Medal shall be asked to present their 
portraits to the Society, and, at the discretion of the Committee, 
the recipients may be asked to prepare a paper for publication in 
the JOURNAL of the Society. These regulations, the names of those 



412 SOCIETY ANNOUNCEMENTS [J. S. M. p. E. 

who have received the medal, the year of each award, and a state- 
ment of the reason for the award shall be published annually in the 
JOURNAL of the Society." 

Active members of the Society are invited, according to the above, 
to propose names of those deemed worthy of receiving the Progress 
Medal Award, which proposals should be seconded by another Ac- 
tive member and forwarded to the Chairman of the Committee, 
Dr. C. E. K. Mees, addressed to the General Office of the Society. 
A written statement of accomplishments should accompany each 
proposal, which should reach the Chairman not later than April 20th. 

The two committees have this year been amalgamated into a single 
committee known as the "Committee on Journal and Progress Medal 
Awards." 

MEMBERSHIP CERTIFICATE 

Associate members of the Society may obtain the membership 
certificate illustrated below by forwarding a request for the same to 
the General Office of the Society at 33 W. 42nd St., New York, N. Y., 
accompanied by a remittance of one dollar. 



Society )Joti<Hi Picture Engineers 




THIS IS TO CERTIFY THAT 



Society of Motion Picture Engineers 




Mar., 1932] SOCIETY ANNOUNCEMENTS 413 

LAPEL BUTTONS 




There is mailed to each newly elected member, upon his first 
payment of dues, a gold membership button which only members 
of the Society are entitled to wear. This button is shown twice 
actual diameter in the illustration. The letters are of gold on a 
white background. Replacements of this button may be obtained 
from the General Office of the Society at a charge of one dollar. 



SUSTAINING MEMBERS 

Agfa Ansco Corp. 
Bausch & Lomb Optical Co. 

Bell & Howell Co. 

Bell Telephone Laboratories, Inc. 

Carrier Engineering Corp. 

Case Research Laboratory 

Du Pont Film Manufacturing Co. 

Eastman Kodak Co. 

Electrical Research Products, Inc. 

Mole-Richardson, Inc. 

National Carbon Co. 

RCA Photophone, Inc. 

Technicolor Motion Picture Corp. 



BACK NUMBERS OF THE TRANSACTIONS AND JOURNALS 

Prior to January, 1930, the Transactions of the Society were published quar- 
terly. A limited number of these Transactions are still available and will be 
sold at the prices listed below. Those who wish to avail themselves of the op- 
portunity of acquiring these back numbers should do so quickly, as the supply 
will soon be exhausted, especially of the earlier numbers. It will be impossible 
to secure them later on as they will not be reprinted. The cost of all the available 
Transactions totals $46.25. 



No. Price No. 


Price 


1Q17 r 3 $0.25 

917 \ 4 0.25 


1924- 


18 
19 


$2.00 
1.25 


1918 7 0.25 




20 


1.25 


1920 
1921 


10 1.00 
11 1.00 
12 1.00 
13 1.00 


1925 


21 
22 
23 
24 


1.25 
1.25 
1.25 
1.25 


1922 
1923 


14 1.00 
15 1.00 
16 2.00 
17 2.00 


1926' 


25 
26 
27 
28 


1.25 
1.25 
1.25 
1.25 



1927 



1928 



1929 ^ 



No. 

29 
30 
31 
32 
33 
34 
35 
36 
37 
38 



Price 

$1.25 



1.25 
1.25 
1.25 
2.50 
2.50 
2.50 
2.50 
3.00 
3.00 



Beginning with the January, 1930, issue, the JOURNAL of the Society has been 
issued monthly, in two volumes per year, of six issues each. Back numbers of all 
issues are available at the price of $1.50 each, a complete yearly issue totalling 
$18.00. Single copies of the current issue may be obtained for $1.50 each. 
Orders for back numbers of Transactions and JOURNALS should be placed through 
the General Office of the Society, 33 West 42nd Street, New York, N. Y., and 
should be accompanied by check or money-order. 
414 



JOURNAL 

OF THE SOCIETY OF 

MOTION PICTURE ENGINEERS 

SYLVAN HARRIS, EDITOR 
Volume XVIII APRIL, 1932 Number 4 



CONTENTS 

Page 

The Problem of Projecting Motion Pictures in Relief 

H. E. IVES 417 

The European Film Market Then and Now 

C. J. NORTH AND N. D. GOLDEN 442 

Victrolac Motion Picture Records F. C. BARTON 452 

Optics of Projectors for 16 Mm. Film A. A. COOK 461 

Silica Gel Air Conditioning for Film Processing 

E. C. HOLDEN 471 

Measurements with a Reverberation Meter 

V. L. CHRISLER AND W. F. SNYDER 479 

16 Mm. Sound Film Dimensions R. P. MAY 488 

Proposed Change in the Present Standards of 35 Mm. Film 

Perforations A. S. HOWELL AND J. A. DUBRAY 503 

The Animatophone A New Type 16 Mm. Synchronous Disk 

Reproducer A. F. VICTOR 512 

The Acoustics of Large Auditoriums S. K. WOLF 517 

Committee Activities: 

Report of the Sound Committee } 526 

Abstracts .. 530 

Patent Abstracts 533 

Book Reviews 536 

Officers 537 

Society Announcements 538 

Spring Convention, Arrangements Program 540 



JOURNAL 

OF THE SOCIETY OF 

MOTION PICTURE ENGINEERS 

SYLVAN HARRIS, EDITOR 

Board of Editors 

J. I. CRABTREE, Chairman 

L. DE FOREST A. C. HARDY F. F. RENWICK 

O. M. GLUNT E. LEHMANN P. E. SABINE 



Published monthly at Easton, Pa., by the Society of Motion Picture Engineers. 

Publication Office, 20th & Northampton Sts., Easton, Pa. 
General and Editorial Office, 33 West 42nd St., New York, N. Y. 



Copyrighted, 1932, by the Society of Motion Picture Engineers, Inc. 



Subscription to non-members, $12.00 per annum; to members, $9.00 per annum, 
included in their annual membership dues; single copies, $1.50. A discount 
on subscriptions or single copies of 15 per cent is allowed to accredited agencies. 
Order from the Society of Motion Picture Engineers, Inc., 20th and Northampton 
Sts., Easton, Pa., or 33 W. 42nd St., New York, N. Y. 

Papers appearing in this Journal may be reprinted, abstracted, or abridged 
provided credit is given to the Journal of the Society of Motion Picture Engineers 
and to the author, or authors, of the papers in question. 

The Society is not responsible for statements made by authors. 

Entered as second class matter January 15, 1930, at the Post Office at Easton 
Pa., under the Act of March 3, 1879. 



THE PROBLEM OF PROJECTING MOTION PICTURES 

IN RELIEF* 



HERBERT E. IVES** 



Summary. The essential conditions for producing pictures in stereoscopic 
relief are two: First, separate pictures must be made from different points of view, 
corresponding to the two eyes; second, each eye of the observer must receive its appro- 
priate view. No compromise with these fundamental requirements appears possible. 

If stereoscopic projection is to be achieved in such a form that a large group of 
observers may simultaneously see the projected picture in relief, the distribution of 
the appropriate views to the two eyes must be accomplished for each observer. There 
are two places where the distribution may be made: the first is at the observers' eyes; 
the second is at the screen on which the picture is projected. 

If the first method be employed, two separate images must be provided on the screen, 
and every observer must have means for directing one image to the right eye and one 
to the left eye. 

If distribution of the images is to be made at the screen, two images are no longer 
sufficient. Theoretically an extremely large number must be provided, a separate 
one for each position that can be occupied by any eye in the audience. 

Several methods of utilizing the parallax panoramagram method are discussed. 
It appears that from the theoretical standpoint the problem of relief projection is 
entirely soluble, and experimental tests of still picture projection have been success- 
fully made. Practically, the solution of relief projection of motion pictures will 
depend upon the use of apparatus involving excessive speeds of operation, great 
multiplicity of taking or projecting units, projection screens containing minute 
ridged reflecting or refracting elements of extreme optical perfection, projection lenses 
of extraordinary defining power, microscopic accuracy of film positioning, and photo- 
graphic emulsions of speeds at present unknown. 

The perception of relief in vision, that is, the location of different 
objects in the field of view at their proper relative distances from the 
eyes, is contributed to by a number of factors. We may list among 
these: geometrical perspective, according to which objects decrease in 
angular extent with the distance from the eyes; aerial perspective, by 
which distant objects are more or less veiled by intervening atmos- 
pheric haze; the effort of focusing or accommodating the eyes to 

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. 
** Bell Telephone Laboratories, New York, N. Y. 

417 



418 HERBERT E. IVES [J. S. M. P. E. 

objects at different distances; and, when the observer can move, by 
the different relative angular motion of near and distant objects. All 
these factors have been utilized to stimulate relief by makers of pic- 
tures both still and moving. The most important factor, however, 
and the only one that needs discussion as a problem still awaiting 
practical solution is binocular vision, which is peculiar to man and 
certain of the higher animals, because of the location of the eyes side 
by side, both receiving images of the same objects. I shall, therefore, 
in this discussion, proceed at once to binocular or stereoscopic relief, 
and our problem will be to consider the ways and means by which 
motion pictures might be projected so as to exhibit relief of this 
character. 

BINOCULAR OR STEREOSCOPIC RELIEF 

While the complete explanation of the process by which we ap- 
preciate relief when the two eyes receive images which are somewhat, 
but not too different, in character, has not been worked out to the 
satisfaction of psychologists, the essential physical conditions of 
stereoscopic relief are simply stated. They are as follows: (1) 
Separate pictures must be available, made from different points of 
view, corresponding to the two views that are seen by the right and 
left eyes. (2) Each eye of the observer must receive its appropriate 
view. These conditions are essential and inescapable. No com- 
promise with them appears possible. No scheme which calls for a 
single picture or series of pictures taken from one point of view will 
meet the first requirement. No scheme which does not provide means 
for distributing the appropriate views to the two eyes will meet the 
second requirement. Once stated, these requirements appear ob- 
vious, and they have indeed been clearly understood by students of 
optics for approximately 100 years. In spite of this, however, would- 
be inventors continue with surprising regularity to announce schemes 
for projection in relief which they claim require no special camera or 
form of picture, or, if they propose taking two pictures in order to meet 
the first requirement, evade the provision of means for separating 
these pictures in the process of viewing. 

Having now cleared the ground, we are prepared for a straight- 
forward discussion of our problem. For purposes of presentation, we 
may conveniently discuss it in three steps : The first step will be the 
production of relief pictures by processes which do not involve pro- 
jection. The second step will take up relief pictures produced by 



April, 1932] MOTION PICTURES IN RELIEF 419 

projection processes, but in the form of "stills," that is, not embodying 
motion. The third step will be to consider the projection of relief 
pictures in motion. 

METHODS OF MAKING RELIEF PICTURES 

In accordance with the requirements as stated above, the first 
piece of special apparatus which is needed in order to produce a 
picture in relief is some form of camera (we shall, of course, assume 
that the process of producing pictures is photographic), which can 
produce pictures from a number of points of view. In the simplest 
case, the number of points of view will be two, one for each eye, the 
apparatus consisting of a pair of similar cameras whose lenses may be 
separated by approximately the distance between the two eyes. 

Pursuing this simplest method of making relief pictures, that is, 
simple stereoscopic pictures of the old and well-known form, we may 
now go over to the viewing end and consider means of meeting the 
second requirement: namely, the distribution of the two pictures to 
the appropriate two eyes. The simplest apparatus for viewing two 
pictures, one at each eye, consists of no apparatus at all, but lies in the 
proper directing of the two eyes. Holding up a pair of stereoscopic 
prints in front of the eyes, with the right eye view at the right and the 
left eye view at the left, one can, by practice, learn to diverge the 
optic axes and see one picture with each eye ; or, if the two pictures 
are mounted side by side, but in the reversed relative positions to 
those just considered, one can, by converging the optic axes to a point 
between the eyes and the pictures, again see one picture with each eye, 
and thus produce a picture in stereoscopic relief. 

Next in order of complexity of viewing device is some form of 
stereoscope. This may consist of mirrors or prisms placed one over 
each eye, and so directed or of such angle as to present one view to 
each eye, the eyes being in their normal unconverged or undiverged 
position. The stereoscope is an instrument very familiar to students 
of optics, and in a previous generation achieved wide popularity as a 
form of entertainment. In our present more feverish age, the appeal 
of pictures without action, even though possessing another aspect of 
naturalness, is so slight that it is now not unusual to find people who 
have never looked through a stereoscope. 

Another means of distributing the pictures to the appropriate eyes 
is provided by utilizing color. In the anaglyph, the two elements of 
the stereoscopic pair are printed in complementary colors, and special 



420 



HERBERT E. IVES 



[J. S. M. P. E. 



spectacles are provided for the observer with a screen of different color 
for each eye, whereby only one picture is seen through either element 
of the spectacles. 

The revolutionary idea that the distribution of the different views to 
the two eyes might be made, not at the eyes of the observer, but at the 
picture itself, was introduced by Frederic E. Ives about thirty years 
ago in the invention of the parallax stereogram. This device, since it 
is the direct ancestor of the most interesting projection methods which 
I shall describe, demands careful description and comprehension. 
According to requirement (1), as stated above, two pictures are taken, 




FIG. 1. The principle of the parallax stereogram. 

from two points of view. Instead, however, of being mounted side by 
side as in the ordinary stereogram, these pictures are divided into 
very narrow strips, these strips being juxtaposed so that the left-hand 
strip of a pair is from the right eye view, and the right-hand strip from 
the left eye view. Close to this picture of alternate strips, which is in 
the form of a transparency, is mounted an opaque line grating with its 
clear spaces approximately half the width of its opaque spaces. This 
grating is mounted at such a distance in front of the stripped picture 
and in such relative lateral positioning of its lines that at a certain 
distance from the observer's face, the right eye strips are entirely con- 



April, 1932] 



MOTION PICTURES IN RELIEF 



421 



cealed from the left eye and the left eye strips are entirely concealed 
from the right eye. Each eye then sees only a single view composed of 
a series of strips which, however, are made of such fineness (say, 100 to 
the inch) as to be invisible or unobjectionable at the viewing distance. 
This parallax stereogram, when held directly in front of the face, 
parallel to the two eyes and at the proper distance, exhibits stereo- 
scopic relief without the interposition of any viewing device located at 
the observer's eyes. The principle of the parallax stereogram is 
illustrated in Fig. 1, and Fig. 2 is a photomicrograph of a small portion 
of an actual parallax stereogram transparency. 




FIG. 2. Photomicrograph of portion of parallax stereogram showing alter- 
nating juxtaposed strips from right and left eye images. 

A limitation of the parallax stereogram is that it must be viewed 
from a single definite direction and distance. While this detracts but 
little from the appeal of the picture if only one observer is to be con- 
sidered, it is a serious defect if, as must be the case when we come to 
discuss means for projecting pictures visible to an audience, a large 
number of people, variously placed, must observe the relief picture 
simultaneously. In order to achieve a relief picture which shall be 
visible at any distance from any direction of observation, it is neces- 
sary to break away from the idea that stereoscopic relief is essentially 
a matter of two images. Consider that the picture is to be viewed not 



422 



HERBERT E. IVES 



[J. S. M. P. E. 



by one person in one position, but by any number of people in any 
possible positions. It is obvious at once that while each of these 
observers needs only two images to satisfy his two eyes, the total 
number of eyes to be satisfied may be very great. This demands at 
the taking end that some camera arrangement be adopted which will 
make the pictures from a very large number of points of view. At the 
receiving end it demands that the grating, or its equivalent, have 
relatively extremely narrow clear spaces so that, as an observer's eye 
takes up different angular positions, an entirely new composite view 
will be seen. In short, in place of the two strips which are behind 
each grating of the stereogram, there must be an extremely large 
number of minute strips behind each very narrow grating opening, 
and since these strips are (in the horizontal direction) little panoramas > 
I have proposed the name of "parallax panoramagram" for this kind 




L ' L 'R 

FIG. 3. The principle of the paral- 
lax panoramagram. 

of picture which shall exhibit relief from any angle or direction of 
observation. The principle of the parallax panoramagram is illus- 
trated in Fig. 3. Fig. 4 shows, greatly enlarged, a portion of a 
parallax panoramagram positive suitable for viewing through a grat- 
ing with very narrow clear spaces. 

It is evident that the problem of making parallax panoramagrams 
with their large number of points of view, must inevitably call for 
bulky or complicated apparatus. Several methods have been pro- 
posed. The most obvious is to provide a battery of cameras, ar- 
ranged, say, in an arc about the object, with their lenses in close 
juxtaposition. If these cameras are then subsequently used as 
projectors for the pictures made in them, and are all directed to a 
sensitive plate placed behind a grating having very narrow clear 
spaces, the resultant photographic print will, with its grating, consti- 



April, 1932] 



MOTION PICTURES IN RELIEF 



423 



tute a parallax panoramagram. In order to avoid the very large 
number of cameras and printing projectors required by this ele- 
mentary scheme, the alternative has been proposed of using a motion 
picture camera which is moved about the object at a slow rate, while 
the requisite large number of views are taken in succession upon 
a motion picture film. Upon projecting the developed film from a 
projector similarly moved, on a sensitive plate behind a grating, a 
parallax panoramagram is obtained with considerable simplification 
of apparatus, but at the cost of the greater time required for the 




FIG. 4. Photomicrograph of portion of parallax panoramagram showing 
panoramic strips which are placed opposite the narrow spaces of the viewing 
grating. 

successive as contrasted with the simultaneous exposures of the first 
scheme. 

Another method of making parallax panoramagram negatives 
consists once more of a moving camera, but uses a grating in front of 
the sensitive plate and develops the minute panoramas behind the 
grating as the camera is moved relatively to the object, either by 
moving the grating during exposure by the width of its spacing (a 
method due to C. W. Kanolt) or by separating the grating and plate, 
and depending on the sweeping of the beam of light through the grat- 
ing slit across the plate behind it as the relative positions of lens, 



424 



HERBERT E. IVES 



[J. S. M. P. E. 



grating, and plate are altered during the exposure. This method, like 
the one using a motion picture camera, requires a sufficient time for 
exposure for the camera to be moved through an arc or other suitable 
path about the object. 

A third, optically ideally simple, method of making parallax 
panoramagram negatives consists in using a single very large diameter 
lens or concave mirror for providing the different points of view. 
This method requires that the lens or mirror subtend an angle from 
the object as large as it is desired that the final picture be visible in 
relief. For an angle of 60 degrees this requires that the lens or mirror 
have a diameter as great as the distance from which the object is 
photographed. Practically, in order to obtain such angles as this, a 
concave mirror is the only feasible device. An arrangement which 





FIG. 5. 



Method of using a large concave mirror for making parallax pano- 
ramagrams. 



has been used successfully for this purpose is shown in Fig. 5. It 
consists of a strip from a 4-foot diameter concave mirror, in front of 
which is placed a half -silvered plane mirror at 45 degrees. The light 
passing from the object to the concave mirror is reflected back to the 
45-degree mirror and then downward to the sensitive plate, which is 
placed slightly behind a grating having clear spaces Yso the width of 
the opaque. Each element of the concave mirror sees the object from 
a different point of view, and reflects an image in a definite direction 
through the grating lines. By using this scheme, a parallax panorama- 
gram negative may.be made at a single exposure. A certain price 
must be paid for such simplification, which is that the perspective 
relations are disturbed; infinitely distant objects are imaged at the 
focus of the mirror, which lies a relatively short distance behind the 
picture plane, thus restricting the method practically to objects near 



April, 1932] 



MOTION PICTURES IN RELIEF 



425 



that plane. This restriction is, however, already present in any 
practical parallax panoramagram since the definition in the panoramic 
strips necessary to differentiate clearly objects far away from the 
picture plane is much beyond that possible by the "pinhole" action of 
the grating spaces. 

Before going on to the question of projection, a few points with 
regard to still relief pictures of the parallax panoramagram type may 
be noted. As above described, the pictures are transparencies viewed 
through an opaque line grating. The form of grating described with 




FIG. 6. Section of parallax panoramagram structure 
suited for viewing by reflected light. The same struc- 
ture is used for several forms of screen for projecting 
parallax panoramagrams. 

its extremely narrow clear spaces is quite wasteful of light. In its 
place may be substituted a grating composed of convex ridges of such 
curvature as accurately to focus parallel rays on the panoramic strips. 
In order to realize the full advantages of such convex ridges, however, 
it is necessary that the strip picture be printed, not on a flat surface, 
but on a series of surfaces which are concave with respect to the 
ridges already considered. This means that the parallax panorama- 
gram should consist of a sheet provided with front and back convex 
ridges, each of different curvatures, as shown in Fig. 6. The curva- 
tures for this purpose are easily computed, and if the technical difficul- 
ties of preparation are overcome, will provide parallax panoramagrams 
which are not wasteful of light, and in which the panoramagrams are 



426 HERBERT E. IVES [J. S. M. P. E. 

visible equally well from all directions of observation. Another point 
to be mentioned in passing is that while only transparencies have been 
considered, the form of picture just described with its ridged structure 
may be made up as a picture for viewing by reflected light, provided 
the photographic emulsion be backed by some white reflecting mate- 
rial, the picture being printed, of course, to low density. Light inci- 
dent on this doubly ridged structure can only come off from any given 
narrow element of a panoramic strip in a certain definite direction, 
thus meeting the essential conditions. 

One further point must be touched upon as presenting an ever- 
present technical problem. In making pictures for the ordinary 
stereoscope, the photographic lenses, of course, invert each element of 
the stereoscopic pair. It is accordingly necessary when stereoscopic 
pictures are made on a single plate, that the prints be cut in two, and 
each separately inverted. If this be not done, the pictures will 
exhibit in the stereoscope, not stereoscopic, but pseudoscopic relief, 
that is, solid objects sink in instead of stand out. Now, in the prepara- 
tion of parallax stereograms and panoramagrams are involved similar 
inverting operations which must be done by some optical inverting 
device. As an illustration, the pictures made by means of a large 
lens or mirror show pseudoscopic relief if the picture is viewed through 
the grating. In order to obtain stereoscopic relief, the expedient is 
adopted in this case of viewing the grating through the picture. In 
every form of taking and viewing device used for parallax pano- 
ramagrams, a close watch must be kept in the inversions due to the 
optical elements, and means must be adopted for assuring that the re- 
lief is stereoscopic instead of pseudoscopic. 

PROJECTION IN RELIEF 

Taking up now the problem of projecting pictures in relief, the 
logical order is first to study projection of still pictures, leaving until 
the end a discussion of the peculiar difficulties introduced by motion. 
In general, all the methods which we have discussed for producing 
relief pictures are available, with certain modifications for projection. 
The essential feature of projection is, of course, that in place of a 
picture fixed in the plane which is observed, the actual picture used is 
placed in a lantern or other projecting device, and an image, usually 
enlarged, is thrown upon the observing plane, which for convenience 
may be spoken of as the screen. 

Following the same outline as that used in the previous section, we 



April, 1932] MOTION PICTURES IN RELIEF 427 

note, first of all, that the simplest method of projecting pictures in 
relief is to throw upon the screen the two elements of a stereoscopic 
pair, and to look at them directly without interposing an optical 
instrument, diverging or converging the optic axes so that each eye 
appreciates only one picture. All that is necessary, therefore, to 
achieve projection in relief is to project pairs of pictures, and to train 
our audiences to control their optic axes by making themselves 
temporarily cross-eyed, or the reverse, during the projection period. 
While this method of stereoscopic projection is entirely feasible for an 
audience of optical experts who have had a little training and prac- 
tice, it does not appear promising for popular use. 

Proceeding next to apparatus to be placed before the eyes of each 
observer, we note that each person in the audience may wear the 
equivalent of a stereoscope of either the mirror or prism form. Next 
in order is the anaglyph scheme, in which the two pictures are pro- 
jected in different colors, and each member of the audience wears 
colored spectacles. This scheme has been used with success in numer- 
ous demonstrations; it suffers from the limitation that it is not 
applicable to projection in natural colors. Two other schemes, which 
might conceivably be used for non-projected pictures, are nevertheless 
specially feasible with projection and are to be ranked among the 
practical methods of this sort. These are, respectively, projection of 
the two images with polarized light, and projection of the two 
images in quick alternation. In the first of these methods, the 
two images are projected by two projectors, one with light 
polarized, say, in the horizontal plane; and the other with light 
polarized in the vertical plane. Each observer is then provided with a 
pair of polarizing prisms, the prisms being mounted in front of the 
eyes, one vertical and the other horizontal, with respect to its plane of 
polarization. By this means, perfect separation of the two images is 
obtained. In the alternate projection method, the two images are 
thrown on the screen alternately in such rapid succession that they 
appear continuous by persistence of vision. In front of each observer's 
eyes are then placed shutters which expose the two eyes alternately, 
operated in such phase that each eye sees its appropriate image as 
projected. This method of relief picture projection has been success- 
fully demonstrated to a full theater audience. 

These methods of relief projection, which call for separate viewing 
apparatus for each member of the audience, are, optically speaking, 
simple and reasonably satisfactory, and are easily adapted to motion 



428 HERBERT E. IVES [J. S. M. P. E. 

pictures. However, the goal of speculation in relief picture projection 
has always been some means of achieving relief without subjecting the 
observers to the inconvenience of special individual spectacles or the 
picture producer to the expense of the multiple viewing apparatus 
demanded. While it is at present doubtful whether schemes which 
provide the distribution of images to the different observers at the 
screen can approach, in simplicity and feasibility, these methods which 
divide the images at the eyes, they are of great optical interest, and 
I shall proceed forthwith to a discussion of them. 

In discussing projection schemes of this general type, I shall adopt 
an order of presentation which is not perhaps logical, but which ties in 
most closely with the results of our study of non-projected relief 
pictures. I shall proceed at once to the problem of projecting parallax 
panoramagrams in their most fully developed form. Let us imagine 
that instead of putting behind the opaque line grating a transparency 
print from a parallax panoramagram negative (made with its pano- 
ramic strips properly oriented to be placed behind the grating), we put 




FIG. 7. Perspective view of glass or celluloid 
rod from which a translucent projection screen can 
be built up for projecting parallax panoramagrams. 

a translucent screen, and that we remove our parallax panoramagram 
print to a projection lantern placed at an appropriate distance behind 
the grating and screen; we then project this parallax panoramagram 
print upon the screen in exquisite focus and in accurate registration as 
to size, position, and inclination of the panoramic strips behind the slits 
of the grating. If this operation can be performed with the requisite 
accuracy, an observer stationed anywhere in front of the grating will 
see a relief picture which will be indistinguishable from the ordinary 
parallax panoramagram. 

The opaque line grating which we have assumed will, of course, be 
very wasteful of light, and in its place it is preferable to use a ridged 
structure such as has already been discussed. In the case of projection 
we are, of course, interested in much larger pictures than, for instance, 
in show window transparencies. In a screen several feet across con- 
taining 200 or 300 ridges, the individual ridges may be as large as a 
quarter-inch in diameter. This relatively large size makes it feasible 



April, 1932] MOTION PICTURES IN RELIEF 429 

to consider building up the screen of separate rods of transparent 
material, such as glass or celluloid. These rod3 will have a cross- 
section consisting of two flat sides, a front surface of one radius of 
curvature, and a back surface of another radius of curvature such that 
all points of the real surface are in the sharp focus of the lens formed 
by the front surface and the body of the rod. This rear surface must 




FIG. 8. Experimental arrangement for projecting parallax panoramagrams 
upon a translucent screen. 

then be given a frosted or other diffusing finish. When a large 
number of these rods are clamped together they form a screen of the 
desired type, on the back of which the parallax panoramagram print 
can be projected. A single rod for such a screen is shown in Fig. 7. 
An experimental screen, built up of 200 rods, of this form is shown in 
Fig. 8, together with the projection lantern used in an experimental 
demonstration of relief projection by this method. 



430 HERBERT E. IVES [J. S. M. P. E. 

Postponing for the present a discussion of how the slide containing 
the several hundred panoramic strip images is to be made, we can 
discuss the practical difficulties which must be faced in projection of 
this sort. Assuming that the picture to be projected is of ordinary 
lantern slide size and that the picture is to be divided into 500 narrow 
panoramic strips, which would correspond to a screen 10 feet across 
with Y^inch rod elements, we must have on our lantern slide some- 
thing like 150 panoramic strips per inch. Each one of these strips 
must be a complete little panorama containing enough sharply denned 
elements to provide separate images for each pair of eyes in an 
audience spread out through at least 60-degrees angular position in 
front of the screen. As a working figure, if we assume 100 differentiable 
strip elements in each panoramic strip (this corresponds to a separate 
view for each eye 20 feet from the screen, lying within 10 feet from the 
center line of the auditorium), we must have a lantern slide in which 
the resolving power is of the order of magnitude of Vi5,ooo of an inch, 
approximating a wavelength of visible light. Proceeding now to the 
projection lens, thia must, of course, give an accurately rectilinear 
image, in order that the panoramic strips on the slide may be ac- 
curately positioned on the back of the projection screen. Next, the 
defining power of this lens must be such that it images the panoramic 
strips on the backs of the screen rods with exquisite fidelity. Proceed- 
ing now to the rod screen, it is obvious that the individual rods must be 
figured with an accuracy comparable with that found in good optical 
lens work if line elements of approximately one-hundredth the width 
of the rod are to be focused from the back diffusing surface into parallel 
beams to be passed into the observing space. It may be mentioned in 
passing that in place of the transmission screen which has been dis- 
cussed, forms of reflecting screen are also possible in which concave or 
convex cylindrical rods are used. In every case, however, the require- 
ments as to extraordinary perfection of all the optical parts obtain. 

From this rough discussion of the requirements, it is obvious that 
the projection of a parallax panoramagram by this method calls for 
most extraordinary refinement of all the elements concerned. On a 
crude scale, however, it has been found by experiment that the 
procedure can be carried through, and relief projection has been 
accomplished experimentally in this way. 

Taking up now the problem of how to produce the "lantern slides" 
for projection by this method, it may be said in general that any of the 
methods which have been described, such as those employing multiple 



April, 1932] MOTION PICTURES IN RELIEF 431 

lenses, moving lenses, and so on, may be used. However, looking 
ahead toward a procedure which might be applicable to motion 
pictures, the most desirable method would be one in which the pictures 
are made by a single exposure on a single plate. The one method 
which is now available for this is to use a large concave mirror as al- 
ready described. When it comes to making pictures for projection, 
however, a complication is introduced, which is, briefly, that the 
mirror method produces pictures which are too large for insertion 
into an ordinary projection lantern. Due to the physical impossi- 




FIG. 9. Arrangement for produc- 
ing parallax panoramagram nega- 
tives of convenient size. 

bility of producing a lens or mirror which shall be both of such large 
size as to subtend a large angle with ordinary objects, and at the same 
time of such short focus as to produce images as small as a lantern slide, 
the mirror method is, generally speaking, only successful in making 
pictures of natural size. Thus, for making portraits, a mirror having 
a radius of curvature of four feet, with the face and the sensitive plate 
each placed four feet from the mirror, is a practical arrangement. If 
larger or more distant objects are to be photographed, the size of the 
mirror must increase in proportion, as well as the size of the picture 
which is obtained. 



432 



HERBERT E. IVES 



[J. S. M. P. E. 



To overcome this difficulty, the parallax panoramagram negative 
made with a large mirror must be reduced in size, by some photographic 
procedure. The preferred way to do this is to re-photograph the strip 
images, formed behind the grating upon a diffusing glass, directly, in 
the first picture- taking operation. A more satisfactory method of 
obtaining the strip images is to substitute for the grating and diffusing 
glass, a transparent ridged screen. The ridges, in order to assure that 
the final picture shall be stereoscopic instead of pseudoscopic, must be 
of the correct direction of curvature for the kind of projection screen 



Cf) 




FIG. 10. Method of projecting pictures in relief 
using a large number of projections and two gratings 
at the screen. 



which is to be used. For a projection screen of the type we have been 
discussing, the screen in the camera should have concave cylindrical 
ridges, which form minute virtual panoramic images. When a photo- 
graphic lens is placed behind this screen at the proper distance with 
respect to its focal length, as illustrated in Fig. 9, a second reduced 
image is formed which may be made of any desired size, such as that of 
a lantern slide. Prints made from this negative are then suitable for 
projection. It is again obvious that the optical quality of the concave 
ridged screen and of the photographic lens just described must be of 



April, 1932] MOTION PICTURES IN RELIEF . 433 

extraordinary perfection. Also, that the photographic emulsion used 
must be of exceedingly high resolving power. 

The system of relief projection which has just been described is, 
from the strictly scientific standpoint, bearing in mind its limitation to 
objects near the picture plane, a complete solution of the problem of 
projecting still pictures in relief. 

Before going on to discuss the peculiar problems of motion picture 
projection, we may consider some suggestions for evading the severe 
requirements which the ideally complete method just described in- 
volves; in particular, the great practical difficulties of exact registra- 
tion of the projected panoramic strips on the screen elements, and the 
necessity for extraordinary resolving power in the photographic 
emulsion. There are several ways of escaping from these requirements 
which, however, demand giving up the single projector or the single 
image. One method which has been experimentally demonstrated 
consists in projecting images from a battery of projectors. If, for in- 
stance, a translucent screen is mounted with an opaque line grating 
both in front of and in back of it, and a multiplicity of images are 
projected from different directions through the rear grating upon the 
translucent screen, the space in front of the screen will present relief 
pictures from any position or direction. (Fig. 10.) The registration of 
these multiple images upon the screen is a matter of relatively in- 
significant difficulty compared with the registration problem above 
considered. A more practical form of screen is one of the reflection 
type which is exactly similar to the translucent screen above described, 
except that the back surface of the rods is given a diffuse reflecting 
finish as, for instance, with aluminum paint. Such rods have the 
property of reflecting light exactly in the direction from which it is 
incident. A screen built of such rods, therefore, will exhibit to each 
eye in an audience only that picture which originates at the projector 
lying in its line of sight produced backward. With a battery of 
juxtaposed projectors in front of the screen, observers in the space 
between the projectors and the screen see the pictures in relief. An 
experimental apparatus for demonstrating relief projection of this 
sort is shown in Fig. 11. 

Another means of avoiding the accurate registration problem and 
also of avoiding the necessity for a large number of projectors is to use 
a single projector, projecting a rapid succession of images, as from a 
motion picture film, but with the projector arranged to move rapidly 
from side to side through a sufficient distance to sweep through the 



434 



HERBERT E. IVES 



[J. S. M. P. E. 



whole observation angle. This scheme is obviously not very practic- 
able because of the mechanical difficulties involved. It is possible to 
imagine some optical means by which the beam of light from the pro- 
jector could fall upon the screen from different angles without the 
projector itself moving, but these again demand very large rapidly 
moving parts, and are of little promise. 

Still another scheme, which may be described as a hybrid method, 
may be mentioned. Suppose that we again use a single motion 
picture projector, projecting in rapid succession a series of views 




FIG. 11. Experimental apparatus for projecting pictures in relief, using a 
battery of projectors and a ridged reflecting screen. 

which have been taken from different directions as, for instance, with 
a moving lens camera. Suppose that we place immediately behind 
the translucent rod screen, an opaque line grating with very narrow 
clear spaces, and that we move this grating laterally back and forth 
so that each succeeding projected image falls in a series of extremely 
narrow bright lines upon the rear surfaces of the rod elements of the 
screen. (Fig. 12.) If the pictures are projected with sufficient rapid- 
ity, and if the opaque line grating oscillates in exactly the phase rela- 
tions required, we shall, by persistence of vision, again have a projected 



April, 1932] MOTION PICTURES IN RELIEF 435 

motion picture in relief. A modification of this scheme consists in 
removing the opaque line grating from immediately behind the screen 
to the projection lantern, placing it immediately in front of the motion 
picture film, and imaging it accurately upon the back of the rod screen. 
We have here again a problem of very perfect image registration, but 
the problem has to be faced only once with a built-in element instead 
of with every picture. 

There are probably other combinations of apparatus which might be 
devised, but the point which I wish to make is, I think, sufficiently 
clear that our fundamental problem is one of providing a vast 
number of images, and that in order to do this we are inevitably forced 
either to make these images of excessively small size or to resort to a 
multiplicity of apparatus or to a multiplicity of projections in time. 
The whole problem is, philosophically speaking, a manipulation of 



sF. 

FIG. 12. A "hybrid" projection scheme, using a large 
number of images projected in rapid succession. 

space and time elements comparable in many ways with the problems 
presented in television. 

PROJECTION OF MOTION PICTURES IN RELIEF 

Because of what has gone before, the discussion of the specific 
problem of projecting motion pictures in relief can be made quite 
brief. All that is necessary is to take one of the methods which have 
been outlined for still picture projection in relief, and to increase 
its speed to the point where the required number, say, 20, complete 
pictures are projected per second. This calls for all the multi- 
plicity of apparatus, all the accuracy of the constituent parts and 
other features which have been discussed, together with the additional 
difficulties of obtaining greater sensitiveness of the photographic ma- 
terials and of performing the accurate registration operations at high 
speed. 

The specific case of the most scientifically complete method may be 
gone into in some detail in order to illustrate what these requirements 



436 HERBERT E. IVES [J. S. M. P. E. 

amount to. Let us assume that the original film is to be made by 
means of the large concave mirror in conjunction with the transparent 
ridged screen and photographic lens. (Fig. 9.) With this apparatus 
as set up for making still pictures for projection of lantern slide size, a 
ridged screen of approximately 200 ridges has been found by experi- 
ment to photograph down to lantern slide size with some success, pro- 
vided the objects to be photographed lie very closely in the plane of the 
picture. With an intense but practicable illumination, the exposure 
necessary with this apparatus is about one minute, due to the small 
photographic aperture of the mirror and the loss of light occasioned 
by the semi-transparent mirror used for throwing the image to one 
side. For the process to be applicable to motion pictures at 20 frames 
per second, the speed of the photographic emulsion would have to be 
increased by approximately a thousand times. If the picture were 
photographed down to ordinary motion picture frame size, the resolv- 
ing power of the film would, without going into figures, have to be far 
better than anything now available; and, if the number of panoramic 
strips were increased from 200 to 500 (Y^inch strips on a 10-foot 
screen), the resolving power necessary to present individual views to 
100 eyes in a row at a 20-foot distance would simply be impossible of 
attainment because it would demand film images smaller than the 
wavelength of light. Much larger film than that now used would then 
be another special requirement. 

Going over now to the projection of the picture so obtained, the 
problem of the exact registration of the strip images upon the projec- 
tion screen is made excessively difficult by the motion of the film. 
Each image must, in turn, be so accurately positioned that no waver- 
ing of the picture occurs. This means that the film must not shift 
laterally by as much as one-hundredth of the width of the strip image 
or approximately, for the case just considered, Yao,ooo part of an inch at 
the projector. When, in addition to these requirements, we remember 
that warping, expansion, and contraction of the film material would 
injuriously affect the registration, it is sufficiently obvious that pro- 
jection of motion pictures in relief by this method calls for a perfection 
of apparatus and materials quite beyond anything now in sight. 
The alternative methods which were noted in the last section, while 
avoiding the chief difficulties of registration, call, as already noted in 
the case of still projection, either for a multiplicity of projectors or for 
high speeds of projection, which in the case of motion pictures would 
mean some hundreds of times present projection speeds. 



April, 1932] MOTION PICTURES IN RELIEF 437 

CONCLUSION 

As we have reviewed the problem of projecting pictures in relief, it 
appears that there are two clearly differentiated methods: 

(1) Involves the distribution of the images to the observers' eyes by 
means of apparatus individual to each observer. 

(2) Calls for means of producing this distribution at the projection 
screen. 

The practical disadvantage of the first scheme is that it involves 
multiplication of viewing apparatus, and some effort and inconve- 
nience on the part of the observers. The disadvantages of the second 
method, as they have appeared from this analysis, are the excessive 
refinement of all the apparatus parts, which could be avoided only in 
part by having recourse to a multiplicity of projecting units or 
excessive speeds of projection. In their present experimental state of 
development, the special screens and other devices called for by the 
second method of projection are too crude for projecting pictures 
visible to audiences of any great size, and the relief images can be 
produced with any satisfactory degree of definition only if the objects 
of interest lie close to the plane of the screen. This latter objection 
is entirely lacking in the first method of projection. In fact, it may 
be said that in the present state of art, the only good quality stereo- 
scopic projection which is now possible is accomplished by means 
of alternate projection, complementary color filters, polarizing devices, 
or other means operating at the eyes of the observers. The means 
involving distribution of the images at the screen are of great optical 
interest, and may be said to be completely postulated theoretically, 
but their practical realization on anything like a commercial basis 
appears remote. 

It has been tacitly assumed throughout this discussion that, if the 
various projection schemes were worked out to perfection, the 
resultant relief motion pictures would possess qualities of naturalness 
which would add to the appeal of the motion picture. There is, how- 
ever, one general consideration which must be recognized: namely, 
that it is not, speaking broadly, possible to project a picture in relief 
which will be "correct," and at the same time exhibit noteworthy 
relief to all members of an audience of any size, stationed at greatly 
different distances from the screen. Striking stereoscopic relief is 
observed in real life only for relatively close objects, and the amount 
of relief varies with the distance of the observer from the object. If, 
therefore, a scene were projected in relief to natural size in the average 



438 HERBERT E. IVES [J. S. M. P. E. 

auditorium using the "parallax panoramagram" method, in which the 
relief changes as it should with the distance of observers, only those 
members of the audience who were in the front rows would find the 
relief quality much of an addition to the picture. If the first method 
of projection be used, in which only two pictures are taken, it is true 
that all members of a large audience will perceive striking relief, 
since the two eyes of each observer will see definitely different 
pictures; but it is at only one observing distance, namely, that from 
which the original object was photographed, that the relief will be 
correct. At other distances, the two pictures will correspond to 
points of view greater or smaller than the normal distance between the 
eyes, giving exaggerated or diminished relief. 

In ordinary projection, in particular motion picture projection, 
objects are rarely reproduced in their natural sizes; usually the 
screen picture is very greatly magnified. In relief projection, 
magnification presents a difficult problem. In the absence of relief, 
gigantic close-ups produce little or no impression of unnaturalness. If, 
however, a typical close-up were presented in relief, the appearance of 
the picture would inevitably be strange and unnatural to many in the 
audience. For instance, if the relief picture be produced by one of 
the first methods, involving the projection of two images to be sepa- 
rated at the eyes of the observers, all the observers, as just noted, will 
have the same two points of view, which will correspond to eyes 
separated by various distances, according to the viewing position. 
The observers from nearby, for whom the pictured object subtends a 
much larger angle than normal, will be virtually seeing the object as 
though their eyes were separated by several feet. In the case of the 
second kind of relief projection, enlarged images are, strictly speaking, 
ruled out. A magnified image will actually appear magnified; a 
face, for instance, will appear as a giant's face, larger than natural, and 
exhibiting the decreased stereoscopic relief that a large object does as 
compared with a small one of similar shape. (To put it another way, 
if the screen image be magnified, the separation of the eyes of the 
observers should be increased in the same proportion.) Close-ups for 
this kind of projection should be shown in natural size, but should be 
so photographed as to appear located in space in front of the screen at 
such a distance from the observer as to give the desired degree of 
intimacy. This introduces the interesting complication in this kind 
of projection that observers nearer than the point where the image is 
formed in space will be between the image and the screen, and will 



April, 1932] MOTION PICTURES IN RELIEF 439 

get no picture. Practically, it means that no image in space should be 
very far in front of the screen. 

I do not purpose at this time to enter into a detailed discussion of 
these complications, but merely to draw attention to the fact that the 
attainment of entirely correct relief projection would carry with it an 
inevitable restriction in the size of the audience which would get much 
benefit from the added factor of relief. If the relief effects are to be 
entirely natural, the motion picture would have to return to a close 
simulation of the dimensions of the regular stage, abandoning one of 
its unique advances over the stage, namely, the "close-up." Doubt- 
less, were relief projection to become feasible and commonplace, a 
special art would be developed, which would strike some workable 
compromise between the appealing qualities of relief and the unnatural 
distortions which great magnification would introduce. 

REFERENCES 

IVES, H. E.: "A Camera for Making Parallax Panorarnagrams," Jour. Opt. 
Soc. of America, 17 (Dec., 1928), No. 6, p. 435. 

IVES, H. E.: "Motion Pictures in Relief," Ibid., 18 (Feb., 1929), No. 2, 
p. 118. 

IVES, H. E.: "Parallax Panoramagrams with a Large Diameter Lens," Ibid., 
20 (June, 1930), No. 6, p. 332. 

IVES, H. E.: "Parallax Panoramagrams for Viewing by Reflected Light," 
Ibid., 20 (Oct., 1930), No. 10, p. 585. 

IVES, H. E.: "Parallax Panoramagrams with a Large Diameter Concave 
Mirror," Ibid., 20 (Nov., 1930), No. 11, p. 597. 

IVES, H. E.: "Reflecting Screens for Relief Picture Projection,' Ibid., 21 
(Feb., 1931), No. 2, p. 109. 

IVES, H. E.: "Optical Properties of a Lippmann Lenticulated Sheet," Ibid., 21 
(March, 1931), No. 3, p. 171. 

IVES, H. E.: "The Projection of Parallax Panoramagrams," Ibid., 21 (July, 
1931), No. 7, p. 397. 

DISCUSSION 

MR. KELLOGG: When the two views of the Capitol at Washington were 
shown, I wondered how far to the side one could go before the actual difference 
in the size of the pictures makes it practically impossible to merge them. 

MR. IVES: I do not know. Stereoscopic relief can be obtained with one of 
the two images in very poor shape. I imagine one could tolerate a lot of dis- 
tortion and yet get the effect. 

MR. GAGE: These parallax panoramagrams, particularly those that are 
colored, are very pleasing. However, I wonder, although they may be inter- 
esting as novelties, whether they would be desirable as a regular thing in the 
theater. This is a question that ought to be considered here. 

PAST-PRESIDENT CRABTREE : I believe that the mind can imagine a lot of things. 



440 HERBERT E. IVES [J. S. M. P. E. 

If the necessary willingness to believe is created in the observers, they will un- 
doubtedly imagine a certain amount of relief, both in the sound and in the 
picture. 

MR. KURLANDER i In view of the many difficulties in getting true stereoscopic 
effects, is there not a simpler method of creating a pseudo effect that would be 
better than the present flat effect? 

MR. IVES: A great deal can be done by lighting, and by poor depth of focus. 
And where the nature of the subject will permit, the relative motion of the 
camera and object provides a beautiful relief. But this is a subject to which 
I have not given much attention. I was talking about binocular or stereoscopic 
relief. 

MR. VICTOR: May I offer, as my personal belief, that we will eventually 
find the solution to stereoscopic relief in colors? Every artist knows the value 
of color perspective, and I think that some day we shall have a color projection 
system that will give us very nearly the effect of stereoscopic pictures. 

MR. IVES: There is a beautiful painting in the Gardiner Museum, in Boston, 
which I would cordially recommend to every one interested in this line of specu- 
lation. It is a picture by Sargent a group of dancers with musicians in the 
background. It is lighted with spotlights, in a rather long room, and as one 
enters through a door at the back, he obtains a remarkable illusion of relief. 

MR. KELLOGG: Concerning the true stereoscopic effect, when two pictures 
are merged, there is only one plane in registration. Everything else is out of 
registration, and in so far as the images of the two observing eyes fuse in the 
consciousness, there must be a blur. Now, a number of efforts have been made 
to produce three dimensional pictures, I believe, by printing two photographs 
or otherwise combining them into a single picture which both eyes see, and in 
which everything in one plane is shjarp and everything in any other plane is 
double. I should like to know whether, as long as one is willing to center his 
attention on that one plane, it provides any better sense of depth than can be 
gotten from any ordinary sharp picture. 

MR. IVES: My comment on that would simply be that it does not conform 
to the second of my series of points that appropriate images are not distributed 
to the appropriate eyes. 

MR. GREGORY: Has cylindrically lenticulated film been used for the taking 
of parallax panoramagrams? 

MR. IVES: Yes, it has. It is more efficient in utilizing the light. 

MR. MAXFIELD: There is an effect I have noticed quite frequently on en- 
tering a motion picture theater ; someone coming out swings the door open, and 
I see a close-up on the screen standing out in beautiful third dimension. On 
measuring approximately the distance from the place where I had noticed that 
effect, to the screen, it looked as though the close-ups had been taken with ap- 
proximately a four-inch lens. I was viewing the picture from my position, at 
the same angle subtended by the lens when the scene was photographed, assum- 
ing that a long focus lens had been used on the close-up. I should like to ask 
Mr. Ives if he has any information regarding the relative importance of really 
correct perspective versus binocular, where one views the picture from a rela- 
tively long distance. 



April, 1932] MOTION PICTURES IN RELIEF 441 

MR. IVES: I do not know anything about viewing from a long distance. 
It is well known in ordinary photography that for a picture to present correct 
perspective it should be viewed from the distance which held between the lens 
and the plate. And by looking at such a picture at that distance with one eye, 
I have heard that one gets a sensation of relief. What happens is, that he does 
not miss the other eye so much. Personally I do not get a sensation of relief 
by looking at a picture with one eye at that distance. But people tell me they 
do. 

MR. MAXFIELD: I do. 



THE EUROPEAN FILM MARKET THEN AND NOW 
C. J. NORTH AND N. D. GOLDEN** 



Summary. European producers will offer some 450 feature pictures during 
1931. Leading producing countries are Germany, offering over 150 German dialog 
pictures, England some 140 sound features, and France over 100 Frenchdia log films 
for the year 1931. Europe is rapidly wiring its theaters, as indicated in the 10,400 
wired theaters during 1931 as compared with the 4950 theaters wired during 1930, 
over a 100 per cent increase in the short space of one year. 

Elimination of legislation detrimental to American interests occurred in France 
during 1931, and a tightening up of quota legislation was continued in Germany. 
Agitation to increase the quota percentage to 50 per cent in the United Kingdom 
gained very little headway, while other countries that have become picture conscious 
are trying to encourage their own production through subsidies, contingents, or taxes, 
as the case may be. Coupled with the problem of European production competition 
and the artificial trade barriers set up by European governments, is that of supplying 
European countries with dialog pictures in their native tongue. 

While the above obstacles appear difficult to surmount at the present time, the 
ingenuity of American technicians and producers will find a way to solve these 
problems to the extent of producing a sufficient quantity of foreign dialog films at a 
cost that will bring a fair return on their investment. 

The financial destinies of most of our film companies certainly 
all the leading ones come unpleasantly close to depending on the 
state of our film revenues from abroad, of which Europe supplies 
nearly 70 per cent. Back in the days of the silent film, approximately 
30 or 40 per cent of our entire rentals came from overseas. After 
foreign audiences stopped going to see pictures merely because they 
had sound in other words, in 1930 our foreign revenues dropped to 
about 25 per cent. As of today, many authorities consider that they 
have fallen below 20 per cent. Obviously the slack must be taken 
up somewhere, and it is therefore no coincidence that the economies 
forced on the various motion picture companies this year have oc- 
curred just at the time of the curtailment of our foreign revenues. 

Perhaps the most effective way in which to apprehend the condi- 

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. 
** Chief and Assistant Chief, Motion Picture Division, Bureau Foreign and 
Domestic Commerce, Washington, D. C. 
442 



EUROPEAN FILM MARKET 



443 



tion in Europe, and the changes that have taken place, is to con- 
sider briefly the situation that exists in a few of our major European 
markets, after which we can possibly reconstruct a picture of the 
scene as a whole. 

The United Kingdom our most important customer, not only in 
Europe but in the world began to find itself film-minded during 
1930, and by the end of that year was quite strongly entrenched as a 
competitive factor in its own market. Prior to that time, it suffered 
an overwhelming dependence on American pictures to the extent of 
about 80 per cent, even as late as 1929. However, the final establish- 
ment of talking pictures gave Great Britain a medium for the exploita- 
tion of its fine stage traditions, and although slow to realize how quickly 
sound films would dominate the scene, once under way the momentum 
acquired was fairly great. Thus, British production that accounted 
for less than 90 films in 1929, increased to a total of 135 in 1930, and 
will reach better than 140 by the close of this year. It should be 
noted that of this 140, about 108 are the products of six companies, led 
by British International and Gaumont-British with 35 each, all of 
which companies have well equipped plants, extensive distribution 
facilities, and theater outlets aggregating a capital investment accord- 
ing to unofficial estimates of not far from two hundred millions of 
dollars. 

TABLE I 

Production Schedules of Foreign Producers 



Germany 


France 


England 








1930 


1931 




1930 


1931 




1930 


1931 


Ufa 


18 


25 


X-Y-Z* 


9 


11 


British Int. 






Suedfilm 




12 


Gaumont- 






Pictures 


34 


35 


D. L. S. 


5 


19 


Franco 






Gaumont- 






Emelka 


6 


10 


Film- 






British 






Terra 


9 


10 


Aubert 


11 


10 


Corp. 


9 


35 


Others 


136 


88 


P a t h e - 






Gains- 












Natan 


11 


16 


borough 


25 


12 








Jacques 






British & 












Haik 


10 


7 


Dominion 


16 


12 








B r a u n - 






British 












berger- 






Lion As- 












Richebe 


4 


8 


ssociated 


5 


86 








Others 


31 


55 


X-Y-Z* 


















Miscellane- 


















ous 


46 


26 


TOTAL 


174 


164 


TOTAL 


76 


.107 


TOTAL 


135 


134 



* American Producing Co. 



444 C. J. NORTH AND N. D. GOLDEN [J. S. M. P. E. 

In spite of England's financial difficulties, the past year has been 
relatively prosperous so far as motion picture receipts are concerned. 
For, while theater attendance has declined considerably from the 
novelty days of 1929, the outlets for sound film showings are much 
greater. For instance, at the end of 1930, about 2600 houses were 
wired. As of today's date, this number has increased to about 4100 
out of less than 5000 theaters, a gain of nearly 65 per cent. The only 
difficulty here from the American point of view is that, even though the 
quantity of films imported from the United States is still decidedly in 
the majority and may amount to as much as 70 per cent of the total for 
the year, British films are gradually getting more advantageous play 
dates, and in a striking number of cases are grossing up to 50 per cent 
higher. In other words, the British public is at last getting something 
that pleases them, out of their own studios, and even though there is no 
language barrier as in Continental Europe, many of the present run 
of the British-made product have local features, whether of voice, 
setting, plot, theme, or the like, from which they acquire greater 
audience value than those imported from the United States. 

This is a new situation, and is one that supplies food for thought. 
For now that the snowball has gathered momentum, there is no know- 
ing how great it may become. We may see the time when England 
will produce not only enough for the greater proportion of its own 
needs, but will also supply the bulk of the pictures shown in its 
dominions. This is probably a pessimistic outlook for us, but un- 
doubtedly revenues from this market are due for a steady decline. 

The recent recommendation of the Federation of British Industries 
and of the Trades Union Council, to increase the quota to 50 per cent, 
made little headway. The English government decreed, in 1927, that 
a certain proportion of all the films distributed and shown in England 
must be British made. This proportion last year was 10 per cent for 
the distributor and 7.5 per cent for the exhibitor; and this year is 12.5 
and 10 per cent, respectively. That the advantages of the quota 
outweigh its initial disadvantages through the organization of mush- 
room companies offering quota films is a question. In any event, it 
now seems to be recognized that nature should be allowed to take its 
own course, and that no further attempt should be made to legislate an 
increase in film output. 

When we come to Continental Europe, the language factor im- 
mediately appears, and must constantly be borne in mind in any 
consideration of France and Germany. In the former country great 



April, 1932] EUROPEAN FILM MARKET 445 

strides have been made in production. In 1929, about 52 films were 
produced (mostly silent); in 1930 this was increased to 76 films in 
sound alone, as well as 18 silent films and a number in foreign dialog. 
This number will be further increased this year to 107 sound and 
dialog pictures, exclusive of foreign versions. Of those pictures 
produced, Pathe-Natan and Gaumont-Franco-Film Aubert are the 
leaders; but in France production in general is considerably more 

TABLE II 

EUROPE 
Increase in Number of Wired Theaters, 1930-1931 

October, 1930 October, 1931 

United Kingdom 2600 4100 

Germany 940 2000 

France 350 850 

Sweden 90 550 

Italy 120 450 

Czechoslovakia 75 300 

Spain 145 290 

Austria 55 295 

Denmark 45 200 

Netherlands 95 180 

Hungary 70 175 

Switzerland 65 140 

Rumania 50 135 

Belgium 30 ... 125 

Yugoslavia 35 110 

Poland 60 105 

Finland 15 70 

Greece 20 70 

Norway 30 60 

Turkey 10 40 

Bulgaria 10 35 

Portugal 15 30 

Other European countries 25 90 

TOTAL 4950 10,400 

decentralized than in this country, no less than 42 companies engaging 
in the production of pictures in 1930, of which 27 produced only 1 
film each. 

French films have in general great popularity, some indeed such as 
Rene Glair's Sous les Toite de Paris and Le Million grossing sums be- 
yond what all but a very few American films have grossed in the most 
prosperous era of silent pictures. Here again, French stage tradition 
and the opportunity of hearing the French language spoken by French 



446 C. J. NORTH AND N. D. GOLDEN [J. S. M. P. E. 

actors has brought about a strongly nationalistic attitude on the part 
of French audiences, with the result that American companies must 
supply French language pictures in order to do business in France. 

And yet the French have to contend with one of the most difficult of 
film problems, namely, product shortage and an insufficiency of play 
dates to enable them to secure sufficient revenue on the average run 
films to expand their production to any marked degree. Obviously, 
their own production is insufficient to meet their demands. And 
outside of that, they have only the French versions of American made 
films and an additional few supplied by Germany. Of the former, 
only about 25 were made in Hollywood so far this year, with possibly 
an equal number in Germany. The result was that at the end of last 
June the French abolished their quota system in favor of limitations 
against only those countries which themselves have restrictions. 
Thus, free entry of American product into France is assured. And 
there is no question that the French market can absorb as many films 
from the United States as our companies are likely to put out, the 
understanding being that these must be in the French language. 

I referred a moment or two ago to France's insufficiency of play 
dates. There are about 3000 theaters in France. At the end of 1930 
some 350 theaters were wired, and even now the number is only 850. 
With less than one-third of the theaters in France adapted to sound 
reproduction it can readily be seen how far the French market is from 
realizing its true potentialities. Obviously, the silent exhibitors 
must wire or go to the wall. 

When we turn to Germany we find that the film business is at an 
exceedingly low ebb. The economic depression is having a retarding 
influence on theater attendance ; and many exhibitors are on the verge 
of bankruptcy which has its painful repercussion on both distribution 
and production. In addition, high taxes threaten to take what small 
profits the few exhibitors are making. This situation incidentally is a 
hold-over, more or less, from last year. Germany has, furthermore, 
continued her rigid policy of import restriction against foreign sound 
pictures, only 105 of these being permitted entrance between July 1, 
1931, and June 30, 1932. Incidentally, this official trade barrier has so 
far been more or less meaningless to the American trade, which has 
not produced German dialog films in quantity beyond what they 
could get permits for and it must be remembered that German 
audiences insist on German dialog but it definitely limits our future 
chances of deriving much revenue from Germany in the near future. 



April, 1932] EUROPEAN FlLM MARKET 447 

In spite of the rather pessimistic picture just drawn of German film 
conditions, it must not be understood that no money at all is being 
made in Germany. As a matter of fact, the bankruptcies are most 
numerous among the smaller and weaker elements in the industry. 
Ufa and Emelka, the latter now being reorganized, are the two largest 
producer-distributor-exhibitors in Germany. The former produces 
approximately 25 features a year and spends upward of $80,000 on 
each of them. In addition, it controls 170 theaters, including many 
of the best locations in Germany. The Emelka chain consists of 50 
first-run houses, and produces 10 to 15 films a year. These and two or 
three other companies are said to be making money Ufa just de- 
clared a six per cent dividend and taken together, they will account 
for a large proportion of the 164 films to be produced in Germany dur- 
ing 1931. This, with such foreign versions as may be secured from 
French and American sources, will come fairly close to filling the needs 
of the German market though there is the possibility of a product 
shortage. It is to be noted that German films are designed not only 
for the needs of the domestic market but to compete actively 
throughout all Central Europe where German, even if not the primary 
language, is generally understood. Special agreements have been 
made with Austria, Czechoslovakia, and other countries, by which 
German pictures gain easy entrance. Germany is also concentrating 
on foreign versions, particularly French, and may soon have as many 
as 40 films for that market. The intensity of this competition and the 
headway it has made must be considered by American film interests as 
at least a subsidiary factor in our diminishing film revenues from 
Europe. 

As to German exhibition outlets, at the end of 1930 about 940 
German theaters were wired. This number has now increased to 
about 2000. The equipment used is German made, mostly Tobis- 
Klang film. There still remain more than 3000 theaters not wired 
which exhibit only old silent films and which must be wired or pass 
out of existence. 

The rest of Europe can be covered in a few words. Little is to be 
expected from Italy, where the ban on all foreign dialog films has 
created such an acute product shortage that fewer than 50 features are 
available for a market that requires over 250 a year. Nearly half of 
these are being produced by Pittaluga, also one of the largest ex- 
hibitors, but the product is not sufficient even for his own houses. 
Obviously, American features in Italian dialog must be very cheap 



448 C. J. NORTH AND N. D. GOLDEN [J. S. M. P. E. 

to show a profit, and with English dialog banned, even when explained 
by Italian sub-titles, there will be very little of the film product of this 
country seen on Italian screens. I might add that the Italian situa- 
tion is merely an intensification of what has been going on since the 
early days of talking pictures. 

Central Europe, meaning Austria, Hungary, Czechoslovakia, 
Poland, and the Baltic States, as implied above, is being drawn 
somewhat into the German sphere of picture influence. They have 
become picture conscious, however, and are trying to encourage their 
own production through subsidies, contingents, or taxes, as the case 
may be. The Scandinavian states, and especially Sweden, are also 
trying production in their own language. American films are, of 
course, being shown in all these countries, but the language obstacle is 
difficult to overcome. The outlet for sound pictures is gradually 

TABLE III 
FOREIGN THEATERS 

Approximate Number of Theaters Approximate Number of Theaters 

Wired 

1930 1931 1930 1931 

Europe 27,000 29,535 4,950 10,400 

Far East 4,000 5,350 900 1,900 

Latin America 4,000 4,700 450 1,575 

Canada 1,100 1,100 450 700 

Africa 750 770 40 116 

Near East 50 85 10 25 

TOTAL 36,900 41,540 6,800 14,716 

being extended in all of them through increased wirings, with the 
result that Europe is far more overwhelmingly committed to sound 
pictures than even figures would seem to indicate. 

As a summary, the charts show, on the one hand, European produc- 
tion, and, on the other, the expansion in European play dates through 
increases in wiring. They provide an illuminating picture, espe- 
cially on the production side, with well over 400 films offered in com- 
petition to our own. 

All told, one must remember that the American trade is faced with 
two important obstacles in Europe. The first is the language ques- 
tion and its subsidiary competition, the latter being almost the direct 
result of the former. The various European countries must have 
films they can understand, and until we can devise a method economi- 
cally profitable to give them such films, with the additional factor that 
they must be of a quality to compete with locally produced films 



April, 1932] EUROPEAN FlLM MARKET 449 

molded on native stage traditions, this problem will not be solved. 
In fact, it is doubtful whether the correct solution to this has been 
given either by that school of thought which advocates production 
of foreign versions abroad, or those that believe production can best be 
done at home. Perhaps the newest types of dubbing, if not too costly, 
will come closest of all to the solution, particularly when applied to 
films in which action predominates. In addition, and this is the second 
obstacle, we have to run the gauntlet of contingents, subsidies, and 
other forms of government protection designed to foster the develop- 
ment of the home product. These may tend to decrease when it is 
comprehended that an industry cannot be legislated into existence; 
but at present they, in combination with high taxes, are doing 
much to make the European film field a series of pit-falls for the un- 
wary. 

In order to brighten the picture, it may be well to state that these 
somewhat pessimistic observations on the decline of our European 
revenues do not necessarily imply that these revenues will reach the 
vanishing point. Far from it. This is a period of adjustment. If 
competition is increasing, so also are film outlets through an increase 
in the number of wired theaters. Europe is going through a profound 
depression which is keeping many people out of the theaters, and is 
impeding theater construction. When things pick up, and with 
better theaters, the chance of increased revenue from an individual 
picture will be greater. In other words, we can make more money on 
fewer pictures. And finally it might not be presumptuous to believe 
that the ingenuity of our producers will find a way to solve the 
language difficulty to the extent that we shall be able to turn out 
foreign language films in sufficient quantity and quality, at a cost that 
will bring us a fair return on the investment. The easy-money 
Europe of silent picture days is gone, but as a market offering better 
returns than now, it holds possibilities. 

DISCUSSION 

MR. RICHARDSON: My reports from France indicate that the projection 
of pictures in France, both as regards sound and the picture itself, is nothing less 
than terrible as compared with our own. The same is largely true in Germany. 
And there is no question, gentlemen, but what that very largely decreases the 
revenue of theaters. I believe that the reason why the photoplay theaters in 
North America are so well patronized is that the picture and sound are repro- 
duced by expert men. 

I believe that the producers might well call the attention of European ex- 
hibitors to the fact that they cannot possibly obtain the requisite revenue if 



450 C. J. NORTH AND N. D. GOLDEN [J. S. M. P. E. 

they put on the screen a very poor picture, and radiate from the horns sound of 
very unsatisfactory quality. 

MR. McGuiRE: While Mr. Rubin, chairman of the Projection Practice 
Committee, was in France about a year ago, he reorganized the entire projection 
staff of Publix Theaters in that country. That program included raising the 
compensation of the men and improving their standing. If these methods were 
more generally adopted in foreign countries much better projection would be 
secured. The importance of projection and of the projectionist is now fully 
realized in the United States, and other countries would do well to follow our 
example along these lines. 

PAST-PRESIDENT CRABTREE : In connection with the matter of producing films 
with foreign dialog, in Hollywood I saw a synchronization of the dialog of a 
foreign actress with the lip movements of an American actress. When the 
picture was projected on the screen, Italian actors and actresses equipped with 
ear-phones, were arranged in front of the screen before a number of music stands. 
By watching the screen and listening to the sound coming from the horn, re- 
markable synchronization was effected. This method of synchronizing is beyond 
the experimental stage now, and the films are now being supplied to the trade. 

MR. GOLDEN: It is true that our technicians at the studio have been able 
to produce a fine result by synchronizing the foreign language and the lip mo- 
tions of our American actors. However, there is one obstacle that they have 
not yet been able to overcome, and that is the question of the proper language 
as used in the country for which the version is made. 

In New York the other day I had the pleasure of talking with a man con- 
nected with the foreign department of one of our large producing units, and his 
complaint was that regardless of how short a time a foreigner has been in this 
country, even as short as a six months' period, there is something that creeps 
into his language that is offensive to the native foreigner in his own country. 
The producer, to secure the true speaking language of a given country, must 
bring the cast from their native country and use them for a certain number of 
pictures, release them and send them back to the country from which they came, 
and then bring over other native actors. The synchronization part of it is all 
right, but idioms of expression, and a certain amount of slang, get into the 
foreigner's speech that are not acceptable abroad. 

MR. KELLOGG: How nearly universal is the standard film track location and 
offset? 

MR. GOLDEN: It is practically the same as used in this country. From 
reports, and samples of film we receive from foreign countries, it is practically 
the same. I am quite sure that Klangfilm-Tobis is about the same as Western 
Electric or any one of our recording systems. 

MR. MONOSSON: Does Europe include the U. S. S. R.? 

MR. GOLDEN: No. In Table II Soviet Russia is excluded because this 
country does not maintain diplomatic relations with Russia, and we are in no 
position to receive authentic information from our own offices. Foreign audiences 
insist on our American stars. It is going to be some years before the foreign 
producer can establish his stars to the point where our American stars will be 
rivaled, and since the foreigner likes our American stars, he must like our 



April, 1932] EUROPEAN FILM MARKET 451 

technic in the production of motion pictures. And as long as the foreign pro- 
ducer, therefore, puts out pictures of the type that he is putting out today, he 
is not going to get very far. 

True, most of this production abroad, as a matter of fact, has been sponsored, 
not by the movie goer, but by the business man of the country in question, and 
even reaches so far as the governmental heads. You have the finest example 
in the British quota system. The British quota system was not instituted by 
the producers or exhibitors of the country. It was originally pushed forward 
by the manufacturing interests of England. They felt our American pictures 
were carrying a propaganda into England which resulted in the sale of our 
American goods in England and its dominions. 

The revenues received by the exhibitor abroad are much smaller than those 
received by the exhibitor in this country. He does not get the patronage at the 
box-office that we do. Therefore, he cannot pay the salaries that our American 
projectionists get. He must accept inferior workmanship. 

But, with many of the theaters having an average seating capacity of two 
hundred, passing out of the picture and being supplanted by the de luxe houses, 
I am sure that projection, theater construction, and entertainment values will be 
improved and placed upon a higher plane. 



VICTROLAC MOTION PICTURE RECORDS* 
F. C. BARTON** 

Summary. A new type of disk record, known as the Victrolac record, is described. 
The material of which it is made is a thermoplastic resin, which must be cooled before 
being removed from the mold. The paper discusses briefly the characteristics of this 
material, the time of playing of records made of it, operating features of the tone 
arm and pick-up system, resonance characteristics of the tone arm, and the char- 
acteristics of the chromium needle. 

A new type of disk record has recently been made available to the 
motion picture industry, which record presents a number of ad- 
vantages over previous types in that it has better reproducing quali- 
ties, better wearing properties, is non-inflammable, practically 
unbreakable, water resistant, smaller for equivalent playing time, has 
much lower surface noise, is lighter and flexible. 

The development of the new record has come about through the 
constant search being made by research engineers for new materials 
which would advance the art of record making and would bring about 
an improvement in an art that has remained almost stationary, except 
for minor changes, for a period of twenty years. In the last few years, 
chemical engineers have given a great deal of time and attention to the 
development of synthetic resins, and the outcome of this work has 
resulted in the production of a great many variations of two general 
groups of these resins. The groups comprise those resins which 
polymerize or cure and become infusible and insoluble after the 
application of heat, and those which are thermoplastic but non-curing ; 
in other words, which will flow and mold under heat but which must 
be cooled before removal from the mold. 

A great number from each group have been used experimentally in 
the hope of finding a material which would be modified to give the 
looked-for improvement in record quality and an almost equal 
number of disappointments have been encountered. Approximately 

* Presented at the Spring, 1931, Meeting at Swampscott, Mass 
** RCA Victor Co., Camden, N. J. 
452 



VICTROLAC RECORDS 453 

a year ago, however, a resin of the second group, that seemed to hold 
promise of having the desired characteristics was found. The 
chemical engineers responsible for the development of this resin were 
called into conference with the engineers of the record manufacturers 
and a cooperative program was laid out in which the technics of the 
two groups were combined to further the development of the resin and 
to combine it successfully with other materials, to the end that a satis- 
factory record material might be evolved. A number of months of 
concentrated effort on the part of these two groups of engineers 
resulted in the production of the compound now known as Victrolac. 
This compounded resin has very remarkable properties, and the 
records made from it have many points of superiority over former 
products. 

Among the principal advantages is the greatly reduced inherent 
background or surface noise as compared with former types of record 
material. In the past it has been found necessary to use a large 
groove and to record sounds of great amplitude so that the recorded 
amplitude would be large compared with the amplitude of the surface 
or scratch noise ; and that by this means the music would mask the 
surface noise, or at least make it less noticeable. Advantage has been 
taken of the improved surface conditions of the new material by 
employing a lower amplitude of recording, smaller grooves, and by 
placing the grooves closer together, thus increasing the playing time 
per inch of recorded radius of the record in direct proportion to the 
increased number of grooves per inch, which in this case is from 90 
lines per inch on the old records to 120 or 130 lines per inch on the new 
records. This represents an increase of from 2.7 minutes per inch of 
recorded radius on the old records to 3.9 minutes per inch on the new, 
and since a film 1000 feet long projected at 24 frames per second re- 
quires about 11 minutes to run, the recorded radius of one of the new 
records corresponding to 1000 feet of film will be 2.82 inches. Allow- 
ing y 4 inch f radius or Y 2 inch of diameter for margin and 2.82 inches 
of radius or 5.64 inches of diameter for recording, we have left 5.86 
inches as the center diameter for a 12-inch record which is satisfactory 
as regards frequency response for the width of groove used. The 
decrease in the amplitude of the recording for the case of the smaller 
groove is about the difference between +9 db. for the old records and 
+5 db. for the new records, or the new recording amplitude is about 
60 per cent of the old. The decrease of surface noise is propor- 
tionately much greater than the decrease of recording level. The 



454 



F. C. BARTON 



[J. S. M. P. E. 



surface of the new material is only about 43 per cent of the old, which 
leaves a net gain of approximately 1.4 to 1 in apparent surface noise-to- 
signal ratio in favor of the new material. In other words, if the 
scratch and the recording noises were reduced by equal percentages 
there would be no change in the noise-to-signal ratio, but in this case 
the surface noise level has been reduced much more than the signal 
level; therefore, there is a net gain in performance. 

Fig. 1 illustrates the relative size of the standard groove on the 16- 
inch record as compared with the new groove on the 12-inch record. 
It will be noted that the curvature in the bottom of the groove is the 
same in each case, and that the groove of the new record is merely a 
little narrower and shallower. 



.0065-Hg 




/////////////// 

121 GROOVES PER JNC 






k- .0070-+ 1 




//////////////////y/' 
'-SECTIONAL VIEW OF 9O GROOVES PER INCH, 

////////////////////////////////////////////I/// 



FIG. 1. Relative size of standard groove on 16 -inch 
record, as compared with the new groove on the 12-inch 
record. 



It is well known that the response characteristic, or the ability to 
reproduce from the record certain frequencies, is directly associated 
with both the linear speed and the width of the groove. In other 
words, the higher the linear speed or the narrower the groove, the 
greater is the possibility of reproducing from the record the higher 
frequencies. The narrowing of the groove on the new record ac- 
counts for the fact that smaller center diameters down to 5 x /2 inches 
may be used in the new records, while still maintaining a frequency 
response characteristic equal or superior to that obtained from the 16- 
inch records with the larger center diameter. 

Another advantage of the material used for these records lies in its 
strength and flexibility. On account of these features it has been 
found possible to produce a 12-inch record for motion picture work 



April, 1932] VlCTROLAC RECORDS 455 

weighing approximately 4 ounces, as compared with 24 ounces for the 
16-inch record. In addition to the reduction in weight, the record is 
practically unbreakable. 

These two features make possible a very considerable saving for the 
producer or distributor in shipping the records. Extremely careful 
and cumbersome packing of records is no longer necessary, and ship- 
ments may be made by mail or express without other protection than 
a couple of sheets of corrugated board on either side of the record so as 
to prevent damaging the record surface by allowing it to come into 





Steel Chromium Point 

After playing one 16 -inch shellac record 





Steel Chromium point 

After playing two 12-inch Victrolac records 

FIG. 2. Showing comparative wearing of needles used on 
shellac and Victrolac records. 

contact with other packages. The new record is approximately 
0.040 inch thick. 

Another but possibly less important advantage lies in the decrease 
of abrasion of the needle. An ordinary full-tone steel needle will 
show much less wear after playing one of the new records than after an 
equivalent amount of playing one of the old. (Fig. 2.) 

It would now be of interest to present a few points in which the 
manufacturers of reproducing equipment, and operators of the equip- 
ment, can assist in the full realization of these advantages. The 
inherent strength of the resin itself is relied upon to give the record the 



456 



F. C. BARTON 



[J. S. M. P. E. 



required solidity. This permits using a soft filler which assists in 
reducing the surface noise. None of the hard, highly abrasive fillers 
commonly used in manufacturing records are used. But the strength 
of the material and its ability to withstand abuse do not necessarily go 
hand in hand, and it may therefore be stated that the new material is 
susceptible to injury through improper use. The records have been 
designed to be operated under the same average conditions as the old 
records ; that is, a standard full-tone needle with a pick-up pressure of 
approximately 5 ounces and a needle placement which will bring the 




- CMr# Of TONC ARM 



FIG. 3. Illustrating proper adjust- 
ment of needle on arc tangent to tone 
arm radius 1 inch from center. 



needle within iy 4 inches of the center of the turntable when the tone 
arm is swung to position directly in line with the center. This place- 
ment will make the needle tangent to a circle approximately 1 1 inches 
in diameter when using a tone arm IP/2 inches long. (Fig. 3.) 

A more desirable set of conditions, with particular reference to the 
new record, would first require a pressure on the needle of 3 ounces, a 
pressure which can be maintained by additional counterweighing of 
the reproducing tone arm. Such a simple correction can be made by 
the operator by allowing the pick-up to rest upon the platform of a 



April, 1932] 



VICTROLAC RECORDS 



457 



small postal scale, placing the tone arm in a horizontal position and 
adjusting or adding to the counterweight to get a reading of 3 ounces. 
Second, a displacement of the needle from the center equal to 1 
inch would be required, making it tangent to a circle approximately 
9 3 /4 inches in diameter, which would lie approximately at the middle 
of the recorded area of a 12-inch record. (Fig. 3.) A change of this 
nature may be difficult to make in existing equipment, but if the dis- 
tance between the needle and the center of the turntable is not more 
than I 1 / 4 inches, no difficulty will be experienced. In designing new 




0.2 



1.8 2.0 



FIG. 4. 



0.4 0.6 0.8 1.0 1.2 1.4- 1.6 
A/EEDLE POSIT/ ON BEHIND CENTER P/M 
/N /NCHES 

Relation between needle position and circle of tangency for a tone 
arm Iiy 2 inches long and a tone arm 9 3 /4 inches long. 



equipment, however, this point should be considered, and the place- 
ment should be made so as to get the best results out of the 12-inch 
record. (Fig. 4.) 

The decreased level of the recording which, as I have said, is about 4 
db. below that of the 16-inch record, will make it necessary for the 
operator to increase the gain by a point or two in order to raise the 
volume in the theater up to the level formerly obtained with the 16- 
inch disk. No change will be necessary in the needle, provided a 
normal full-tone needle that is not excessively sharp is used. 

Relatively little consideration has been given in the past to the 



458 F. C. BARTON [J. s. M. P. E. 

shape of the needle point. Although the record material itself has 
been sufficiently abrasive to wear down the needle point rather quickly 
to fit the groove, with the new material this process takes place much 
more slowly; and with the slightly softer record stock, cutting of the 
record may result from either too fine a point or too high a pressure. 
Assume that the combined weight of the pick-up and the tone arm 
is 5 ounces, or roughly 1 / 3 of a pound, and that the area of the 
point in contact with the record is 0.003 inch in each direction, or 
approximately 9 square mils. Under such conditions, if the pick-up 
weighs 1 pound, the pressure under the needle would be 110,000 
pounds per square inch; but since it weighs only l / 3 of a pound, the 
pressure will be of the order of 37,000 pounds per square inch, a fairly 
high stress even for metals. When we consider the nature of the 
record compounds it is remarkable that such a stress can be withstood 
even for a single playing. A reduction in weight from 5 to 3 ounces 
will cause a corresponding reduction of stress from 37,000 to 22,000 
pounds per square inch, a value still quite high for an ordinary thermo- 
plastic molding compound to stand. The existing standard of 5 
ounces was selected to insure tracking of the needle, or following of the 
sound wave, on the very heavily recorded 16-inch picture records; 
but since the amplitude of recording of the new records is considerably 
reduced, there is no longer the need for so great a weight to insure 
tracking, and 3 ounces have been found ample. 

The new records, if used under the conditions recommended above, 
will have a life much longer than any records that have been previously 
produced for the motion picture industry. 

Needle development has been carried on in parallel with record 
development, and there is now available a new type of needle admir- 
ably adapted to the new record, although its use is in no way restricted 
to this record. It is a full- tone steel needle having a chromium tip. 
When used under a 3-ounce load this needle will successfully play at 
least twenty-five of these 12-inch records. A number of playings 
greatly exceeding twenty-five have been successfully made in the 
laboratory, but this number is recommended as representing good 
practice. Assume a 12-reel feature motion picture show running four 
times a day. Twenty-four records would be played on each projector 
each day, requiring a change of needle only once a day per projector. 

Before closing, a short statement referring particularly to the design 
of tone arms and their effect on the performance of a record might be 
appropriate. Judging from the characteristics shown by some of the 



April, 1932] 



VICTROLAC RECORDS 



tone arms that have been tested, their designers apparently have 
considered them as merely means for holding the pick-up in its proper 
position on the record. True, this is one of its functions, but another 
and equally important function is that of controlling the tendency of 
the pick-up as a whole to rotate around the natural longitudinal 
axes of the arm, the impulse causing this tendency being furnished by 
the lateral motion of the needle during the recording. Some tone 
arms, instead of exerting a corrective influence against this tendency, 
by their construction actually tend to aggravate the tendency to 
rotate. The increase of this tendency will, of course, occur at or near 

70N. ARM WITH "U" CROSS SECTION AND LOW 
IMPEDANCE PICK-UP 



600 



500 



400- 



300 



200 



100 



HEEDLE PRESSURE * /7o GR. 




2 4 & Q 100 2 4 6 8 1000 2 4. 

FREQUENCY 
FIG. 5. Resonance characteristic of an undesirable tone arm. 

the frequency at which the tone arm and pick-up would vibrate if they 
were placed under torsional stress and suddenly released; in other 
words, at the period of natural resonance. If this resonance fre- 
quency occurs, as it frequently does, in the lower musical register, 
then a severe load will be imposed on the record; and the needle will 
tend to leave the groove each time the arm is shocked into vibration 
by a passage in the record of a frequency corresponding to the natural 
period of the tone arm system. A curve plotted from data obtained 
from a particularly bad tone arm is here shown in Fig. 5. The con- 
clusion reached from this is that, if it were not possible to design an 
arm free of natural periods, the arm should be designed so that the 



460 F. C. BARTON 

period will occur at a frequency well below 100 cycles, or when the 
recording has been so attenuated that the shocks produced will not be 
large. In general, long straight [/-section channels should be avoided. 
In reviewing the performance possibilities of this record it is the 
firm opinion of the developers and manufacturers of the record that an 
outstanding advance has been made and that with a small amount of 
cooperation by designers and operators of the equipment, the full 
advantages of the new development may be realized. 



OPTICS OF PROJECTORS FOR 16 MM. FILM* 
A. A. COOK** 

Summary. The limits of illumination available in a projector are fixed by three 
factors: the size and brilliance of the light source, the effective aperture of the optical 
system and the design of the condensing lenses. In modern 16 mm. machines of the 
standard type, about 100 to 120 lumens are available through an f/2 optical system; 
these values, which are not corrected for shutter and film losses, mean that 1.6 to 2.0 
per cent of the total radiation is being used. The use of low voltage lamps has not 
changed this ratio to any extent. The effect of varying each of the above factors is 
discussed, and the increase in screen brightness that is likely to be obtained is estimated. 

The fundamental requirements of apparatus designed to project 
motion pictures from 16 mm. film are too well known to need any 
detailed description. The apparatus must be compact and light, and 
the number of adjustments necessary to operate it should be 
reduced to a minimum. As an optical instrument it ought to produce 
a clearly defined image on the screen. It is also obvious that the 
location of the optical elements and their relation to the light source 
must be exactly maintained if maximum illumination is to be con- 
sistently secured. 

Projection optical systems consist of a source of light, a collective 
system for directing the light through the film gate, and an objective 
lens for imaging the film upon the screen. Let us first consider the 
light source. The advantages of tungsten lamps are evident from the 
requirements already outlined. They are small in size, easily located 
in a fixed position, and require a minimum of adjustment during 
operation. Several filament designs of high efficiency have been 
developed with parallel coils arranged to fill a rectangular space about 
two-thirds the size of the film gate opening. The spaces between the 
coils are of approximately the same width as the coil, this arrange- 
ment permitting the use of a spherical mirror behind the lamp to 
image each coil in the adjacent space. This adds to the efficiency by 
heating the filament and gives the unit nearly the appearance of a solid 

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. 
** Bausch & Lomb Optical Co., Rochester, N. Y. 

461 



462 A. A. COOK [J. S. M. P. E. 

source. By doubling the useful angle of radiation in this way an 
increase in illumination of 50 to 75 per cent is obtained. The exact 
amount depends on the quality of the mirror and the position of the 
filament supporting wires. 

The filament housing is a tubular bulb iy 4 inches in diameter. 
This size has been adopted as standard for 16 mm. equipment, al- 
though it may not prove sufficient for the continual demands for 
higher wattage. 1 Bulb diameter is an important dimension from the 
optical point of view. The efficiency of the condenser and reflector 
depend on the angular size of the cone of light that they can take in 
from the source and transmit through the system. A shorter distance 
between filament and condenser would be helpful, therefore, in that it 
would permit a larger angle to be used by a condenser of given di- 
ameter. Lamp manufacturers have been working on this problem, as 
is shown by the fact that in some of their recent designs the filament 
has been offset to a position well forward of the center of the bulb. 
This change provides a mechanical advantage which can be especially 
useful in the 16 mm. projector. Condenser design has often been 
handicapped here by the limited space available. An increase in the 
diameter of the mirror will be necessary, of course, for its distance 
from the filament has been increased. There is more room behind the 
lamp, however, and this slight change can be easily made. 

The collective system may be either a condenser or a reflector. 
Both methods have been applied to the illumination problem in 
projection, but more space is required by a reflector, for the same 
useful angle of radiation, than by a condenser with rear mirror. 
Therefore, the condenser has been the preferred form in 16 mm. 
machines. 

The function of the condenser is a subject that has been thoroughly 
analyzed and presented before this Society. 2 Only an outline will be 
given here of the working of this element of the optical system as it 
applies in this special case. If a solid source of light of sufficient size 
and uniform distribution could be placed at the film gate, no con- 
denser would be needed. A tungsten filament is not solid, however, 
nor can a lamp bulb be placed at that point. By using a condenser 
a source image is substituted for the source itself; by locating the 
image in front of the film plane the unevenness of the source can be 
equalized. Fig. 1 is a sketch showing the condenser in its relation to 
the other parts of the system. The condenser, L\, produces a magnified 
image of the filament of such size as to fill the projection lens, L 2 . In 



April, 1932] 



OPTICS FOR 16 MM. FILM 



463 



doing this it takes in the large angle of radiation marked a, and forms 
the image at a smaller angle a'. The radiation can now be trans- 
mitted through the projection lens L 2 , as a result of this change in its 
direction. In this way the condenser makes useful the radiation 
from a small source through a large solid angle in space. Otherwise, 
a very large source would be needed to produce the same effect. 



SOURCE IMAGE 




PROJECTION LENS 



MIRROR 



FILM GATE 



FIG. 1. Projection optical system for 16 mm. film. 



There is a very definite relation here between the size of the source, 
the size of the projection lens, and the focal length of the condenser. 
All the parts of the optical system are interdependent in this way, and 
proper proportions must be maintained to obtain maximum efficiency 



SOURCE IMAGE 




PROJECTION LENS 
L 2 

FILM GATE 

MIRROR 

FIG. la. 16 Mm. optical system, showing illumination at margin 
of film. 

of the whole unit. The conditions determining the diameter of the 
condensing lenses are shown in Fig. la. Two solid lines drawn from 
the extreme edge of the effective lens opening to the center of the film 
aperture form an angle a'. The broken lines in the same way deter- 
mine angle b' at the margin of the picture. These two solid angles, a' 
and b f , must be equal in size and must be filled with light in order to 



464 



A. A. COOK 



[J. S. M. P. E. 



get the best possible illumination at the corners of the screen. This 
means that the condenser should be large enough to furnish light 
through all of the angle b'. This condition is usually not perfectly 
fulfilled in practice. A 15 per cent decrease of illumination at the 
margin is commonly accepted as satisfactory. 

Condensers constructed according to these specifications are still 
found to differ considerably in efficiency, due to differences in their 
correction for spherical aberration. This is a well-known defect, 
found in all simple lenses, that causes in this instance a loss from the 
marginal portion of the light beam as it is converged to the image point 
by the condenser. The loss is not so serious in 16 mm. projection 
systems as in cases where the source image is located at the film gate. 
It can be corrected to a large extent by proper condenser design. The 




FIG. 2. 



Relay condenser. Conjugate images are connected by 
brackets. 



use of aspheric surfaces is one effective method, this kind of correction 
having been found to result in screen illumination 15 per cent greater 
than that obtained with the ordinary plano-convex condenser lenses. 

The relay condenser is a more complex device that may prove useful 
with 16 mm. equipment. Its use in motion picture work is not new. 3 
But it produces uniform illumination from a tungsten source with so 
little loss that it ought to be included in any discussion that deals with 
projection from filament lamps. As shown in Fig. 2, it is a compound 
lens system composed of three units. There is a condenser system, 
LI, of large angular aperture to image the source upon a relay lens, 
LZ, placed a short distance in back of the film gate. The third element, 
L 3 , serves to form a second image of the source in the projection lens. 
The relay lens must be large enough to receive all of the source image, 
and of such focal length as to form a reduced image of the condenser 



April, 1932] 



OPTICS FOR 16 MM. FILM 



465 



at the film gate. Note that it is the evenly illuminated condenser 
surface, not the source, that is imaged on the film. This accounts for 
the uniform screen illumination produced by the system. It is 40 
per cent more efficient than plano-convex condensers. The extra 
length of the unit, amounting to six inches over all for a 16 mm. outfit, 
is a decided disadvantage. But if it ever becomes necessary to build 
a special type of projector for school or auditorium use, this method of 
illumination should be of great service. It can be constructed to 
work with a small source, and provide sufficient magnification to fill 
larger projection lenses than any that are now used in 16 mm. work. 

The projection objective is the third important part of the optical 
system. Two-inch focus lenses of //2.0 are standard equipment at the 
present time on practically all projectors except those designed for 
use in cabinets. They must be well corrected for this large aperture, 
but the field to be covered is so small that the requirements can be met 



FRONT 



BACK 




FILM 



FIG. 3. Projection objective of Petzval type. 

without difficulty. There are many types of lenses that could be 
used. In any such situation the cost element is bound to be a 
decisive factor, and it has operated in this case to select the least 
expensive lens that can be made to do the work. Before discussing 
the details of this particular lens construction, it would be well to 
consider the original from which it was derived. This lens form, 
shown in Fig. 3, is Petzval's portrait objective. It has undergone 
modification many times, but is still the formula most often used for 
projection work. It can be very precisely corrected for the small 
field required, and has a light transmission, in short focal lengths, of 
73 per cent. 

Fig. 4 shows the modified form that is now used in so many 16 mm. 
projectors. Note that the two rear elements have been cemented, 
and that the spacing between front and back has been increased to 
nearly twice the length of the original construction. The first 
change, by eliminating two air-glass surfaces, increases the light 



466 



A. A. COOK 



[J. S. M. P. E. 



transmission to 81 per cent; the increase in length has the effect of 
shortening the back-focus of the objective. This means that the rear 
element can be made smaller in diameter without sacrificing in light 
transmission, and that it has more space in which to converge the 
beam of light from the film gate. The rear element thus acts as a 
collective lens for the system, which results in the practical advantage 
that objectives of this construction, of any focal length, can be used 
interchangeably on a projector without alteration or adjustment of 
the condensing system. The only disadvantage of this short back- 
focus objective is that it has a slightly curved field. This defect is 
noticeable only in critical tests, however, and would be difficult to 
detect in practical use on a projector, with moving film as a test 
object. 

The final screen illumination produced by a 16 mm. projection 
system depends on the effectiveness of the four elements that have 
been described: the light source, the rear mirror, the condenser 

FRONT 



FILM 
FIG. 4. Projection objective with short back focus. 

system, and the projection objective. Increases can be obtained by 
using a brighter source, by improving the condenser correction, and 
by increasing the aperture ratio of the entire optical unit. Recent at- 
tempts at improvement in the 16 mm. field have been mainly directed 
toward the light source, and this choice is a logical one for the equip- 
ment manufacturer because it involves the least amount of redesign 
on his part. To meet this demand lamps of greater brightness have 
been developed, the increase being due to the use of larger wire size in 
the filaments operated at a lower voltage than previously used. 1 The 
possibilities here are beyond the field of optics, and must be left to the 
electrical engineer. 

There are two points about lamp filaments, however, which are of 
optical interest. One is the fact that filament supporting wires cause 
illumination losses unless they are placed outside the angular field of 
both the condenser and the rear mirror. The second concerns the 
filament itself. The aperture of a projection system must be filled 




April, 1932] OPTICS FOR 16 MM. FILM 467 

with light if it is to work at its best efficiency. With a filament lamp, 
the source acts as a discontinuous surface, and the openings in its 
area cause a real loss of light. This effect is shown in Fig. 5, which 
is a photograph of a 4-coil tungsten filament and its mirror image, as 
they appear at the aperture of a projection lens. Any change that 
would help fill up these spaces and thus make the source more solid 
would mean an increase in illumination. 

Improved condenser design offers a small field for improvement 
which is applicable, perhaps, to many of the 
commercial machines. Even with a perfect 
condenser, however, one can do no more 
than to fill the projections lens with an image 
of the light source. The brightness of the 
source and the effective aperture of the 
system then determine the illumination. 
Increasing the aperture offers interesting 
possibilities that are yet to be considered. FIG. 5. The filament 
An //1. 5 optical system should give 75 per image as it appears in 
cent more light on the screen than the //2.0 the projection lens, 
lenses now used; experience indicates that 

these theoretical increases are seldom attained, however, and that a 
figure of 50 per cent is much nearer the probable increase. The 
cost element enters into this situation to such an extent that an 
increase in aperture is not likely to be attempted in commercial 
practice until all possibilities of the light source have been realized. 

REFERENCES 

1 ROPER, V. J., AND WOOD, H. I.: "Trend of Lamp Development and Opera- 
tion in Motion Picture Projectors Employing 16 Mm. Film," /. Soc. Mot. Pict. 
Eng., 15 (Dec., 1930), No. 6, p. 824. 

2 KELLNER, HERMANN: "The Function of the Condenser in the Projection 
Apparatus," Trans. Soc. Mot. Pict. Eng. (Nov., 1918), No. 7, p. 44. 

3 KELLNER, HERMANN: "Can the Efficiency of the Present Condensing Sys- 
tems Be Increased?" Trans. Soc. Mot. Pict. Eng. (Oct., 1923), No. 17, p. 136. 

DISCUSSION 

MR. PALMER: It has seemed to me, from casual observation, that the propor- 
tions of the filament should be one to three-quarters three-quarters as high 
as it is wide, in order to conform to the dimensions of the picture aperture. Am I 
correct in assuming that? 

MR. COOK: In the case of 16 mm. projectors, we can not get uniform illumina- 
tion when the filament is imaged on the aperture itself. For that reason the 



468 A. A. COOK [J. S. M. P. E. 

image of the filament is moved forward enough to produce the desired effect of 
a uniformly illuminated screen. It actually amounts to imaging the source be- 
tween the projection lens and the aperture. The projection lens is round, and it 
seems to me that a nearly square source would be as valuable as one that is oblong. 
The effect of the aperture in stopping down the light is, of course, noticeable as 
soon as we get the image in front of the aperture. But in order to follow out that 
line of reasoning we should use square condensers and a square projection lens. It 
seems to me that a round source would be more nearly the ideal, from the present 
set-up we are using in sixteen millimeter work. There is no doubt that if we 
image a square source to fill a round projection lens, we waste the light coming 
from the corners of the filament. But the illumination obtained depends on the 
brightness of the source and the effective aperture of the system. 

MR. HICKMAN: It seems to me that it makes no difference how much light 
is spilled over the edge, provided a little more can be obtained in the center. 
No one is really concerned with what is lost around the side. 

MR. KURLANDER: The shape and size of the filament are also governed by the 
desire of the projector manufacturer. Of course, the projector manufacturer 
is susceptible to some influence by the lamp manufacturer, but sometimes he is 
not, and he has his own ideas. I believe that the present trend is toward the 
square shape, the size being dependent upon the inscribed circle determined by 
the back element of the projection lens. Also, with some special forms of optical 
systems, special shapes and sources are required, and those special shapes im- 
mediately give rise to new lamps. Sometimes the new lamps are placed on the 
schedules and are available to other manufacturers who do not know the history 
of their development, and choose from them at random to meet their con- 
ditions. 

So there are a number of reasons for the different shapes of light sources, and 
while theoretically, a solid source should be in the proportion of three to four, 
a square source is easier to construct mechanically and does the same work. 

MR. GAGE: The last picture that Mr. Cook showed was a photograph of the 
projection objective filled with the filament. That is the way it looks when you 
stand at the screen and look at the projection objective through a dark glass, 
while the picture is being projected. If you find that the entire surface of the 
projection objective appears filled with light, when observing the projection 
lens from all points of the screen, the optical system is delivering all the light it is 
capable of delivering. If, on the other hand, you find it is not filled with light, 
perhaps you can tell by simple inspection where the defect lies. Perhaps the 
filament is askew, perhaps the image of the filament in the mirror does not fill 
the space between the filament legs with light, or perhaps the filament is not big 
enough to fill the aperture. If you find a small image of the filament filling only a 
part of the area, a larger filament is required, or perhaps a shorter focus condenser, 
to magnify the filament image to a greater extent. 

If, now, we go through the back-focusing process, setting up the whole pro- 
jection system with the aperture, the objective and so on, and put a light in front 
of the objective, with a card at the focus of the condenser, it can readily be seen 
that there is no use in having a filament any larger than the spot of light received 
on the card. 

One thing Mr. Cook did not explain: if the condenser is brought close to the 



April, 1932] OPTICS FOR 16 MM. FILM 469 

filament, while, at the same time, the surface of the condenser is bent, as can be 
done theoretically, a larger amount of the light will impinge on the first surface of 
the condenser brought into the optical system by simply bringing the same diam- 
eter condenser close to the filament. The acceptance angle becomes greater, and 
the filament image becomes larger, until the surface of the condenser comes into 
contact with the glass bulb surrounding the filament. 

With the present sized filaments, the filament image is sprawled over a larger 
area of the objective than can be used. 

The greatest possibility of improving the system is to increase the intrinsic 
brilliancy of the filament. By using a more efficient condenser it is possible to 
use a smaller filament area. Nothing is gained unless the intrinsic brilliancy 
of the light source is increased. At the same time its area can be reduced by a 
more efficient condenser. 

MR. RAYTON: There is one point that Mr. Cook and Mr. Gage have not 
touched on, that might be worth mentioning since so much attention has been 
paid to the appearance of the front of the objective. If we are to judge whether 
the relative aperture of the optical system is completely filled with light, we will 
have to do something more than look at the front of the objective under normal 
condistions : namely, we will have to insert a pinhole aperture at the center of the 
film gate. We may find, by so doing and we probably will find that the front 
of the objective is filled with light. We will most certainly find, if we move 
the pinhole to the corner of the film gate, that the objective is no longer filled 
with the image of the filament. We may also find cases in which, with the full 
film gate exposed, the front of the objective appears to be filled with the image; 
but that when we introduce the pinhole aperture at the center of the film gate 
the lens aperture is no longer so filled. If it is not, then under those circumstances 
the condenser design is not the most efficient that could be used. 

This point ought not to be overlooked, and I want to emphasize it, that we 
do not get full information about how an optical system is working without an 
exploration carried out with a pinhole aperture at the film gate. 

MR. KURLANDER: I should like to ask if it is not true that the objective lens 
is seldom filled by each individual point of the aperture. 

MR. RAYTON: It is generally true. 

MR. KURLANDER: Then I wonder why so much light is spilled over the 
aperture plate to get uniform screen illumination when it would be cheaper to 
use a cheap lens and a diffusing element in front of the condenser lens. 

MR. RAYTON: Usually because the condenser is not large enough. The 
relative aperture of the projection lens required in order to get center brightness 
is one thing Mr. Cook mentioned the fact that a decrease of brightness of 
fifteen per cent or more. at the margin will pass unnoticed. To get uniform 
quality of illumination all over the screen, we have to use condenser lenses possibly 
somewhat larger than are ordinarily used. 

MR. KURLANDER: Do you have to go to such extremes to get evenness? 

MR. RAYTON: You do with the set-up for motion picture illumination. 

MR. KURLANDER: I have obtained uniform screen illumination of equal 
intensity by focusing the filament at the aperture plane, and then smoothing 
out the light by placing a diffusing element in front of the condenser lens. 

MR. RAYTON: It is quite unreasonable that you should. 



470 A. A. COOK 

MR. KURLANDER: It seems unreasonable, but I hope some time to be able to 
show it. 

MR. COOK: Will back-testing according to Mr. Gage's method in this way 
show that the filament should be round rather than square? 

MR. GAGE: When I tried back- testing the condenser system with the aperture 
in place, I obtained, at the position of the filament, not an exact image of the 
aperture, but approximately that. It is wider than it is high and is rectangular, 
with rounded corners. 

MR. FARNHAM: There is an eternal demand for more and more light from 
projection optical systems. There are three ways of obtaining it: greater 
source brightness, greater efficiency in controlling the light through the optical 
system, and utilizing greater source area without reduction of efficiency. I do 
not see how we can expect an increase of source brightness of a very high order, 
that is, two- or three-fold, as we are now operating the tungsten filaments at 
3400 K., and the melting point of tungsten is 3650, the highest of any metal we 
know. The wire has been so disposed in making a concentrated source as to 
secure an optimum effect of black body radiation and a high order of average 
source brightness to maximum brightness. Further increases will be a few per 
cent at a time. It would appear that the greatest development lies in the direc- 
tion of improved optical efficiency and of utilizing greater source areas. This is 
particularly emphasized when it is realized that the over-all efficiency of the best 
projection optical systems is approximately five per cent. 



SILICA GEL AIR CONDITIONING FOR FILM PROCESSING* 



E. C. HOLDEN** 



Summary. The need of properly conditioning air in motion picture film process- 
ing plants is pointed out and the values of temperature and relative humidity in current 
use in such plants are indicated. In particular, the process of humidifying, in which 
pure silica which has passed through the sol and gel stages of manufacture, commonly 
called silica gel, is used, is described. The principles involved in the process and the 
efficiency of the method in controlling the condition of air are explained, and curves 
are given showing the efficiency of adsorption. 

The most obvious application of air conditioning in the motion 
picture industry, aside from the comfort conditioning of theaters, is 
in film processing. This seems a relatively simple operation, more 
or less satisfactorily performed at a large number of places ; neverthe- 
less, a reactionary attitude of secrecy still prevails, even as to this 
detail, resulting in wide variations in local practice. 

There is probably an ideal set of conditions for film processing, the 
determination of which would be hastened, to the benefit of all, if a 
more enlightened policy of exchange of experience were practiced, 
such as is fostered by this and other technical societies. Cases where 
secrecy in the arts is justified and desirable are the exception rather 
than the rule. 

AIR CONDITIONING REQUIREMENTS 

The usual requirements to be met by air conditioning for health and 
comfort purposes are that : 

(1) The air must have the approximate chemical composition of fresh air. 

(2) It must be free from odors. 

(3) It must be clean. 

(4) It must have an effective temperature within the comfort zone. 

In the case of film processing, the important factors are as given 
under (3) and (4). Processed air should be super-cleaned, as nearly 
free from suspended solids as is possible, and its effective temperature 

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. 
** Consulting Engineer, Baltimore, Md. 

471 



472 E. C. HOLDEN [J. S. M. P. E. 

must be within the "comfort zone" for films instead of for people. It 
must be remembered that the effective temperature with relation to a 
moist surface is determined by the dry-bulb temperature, the relative 
humidity, and the velocity of the contacting air; and that the requisite 
"comfort zone for films" is such that they may be dried rapidly with- 
out suffering distortion or becoming brittle. 

All are agreed as to the desirability of having the air perfectly clean, 
a requirement which has become all the more important with sound 
recording. The standard oil-surfaced baffle and the felt filter types of 
air cleaners are not adequate for this purpose, as the former puts oil 
into the air and the latter lint, both of which may adhere to the film 
and produce highly objectionable effects. 

There is considerable difference of opinion as to the optimum film 
speed, the volume, temperature, and the relative humidity of the air 
to be circulated through the drying cabinets. In practice, tempera- 
tures from 50 to 110F., relative humidities varying from 20 to 80 
per cent, and film speeds ranging from 15 to 130 feet per minute are 
variously used. Even allowing for the difference between positive 
and negative film requirements, such extremes cannot all be right. 

The ideal conditions can be determined only by making systematic 
tests. For this purpose adsorption conditioning units can supply 
clean air at any desired temperature and humidity. 

SILICA GEL 

Silica gel is chemically pure silica, which has passed through 
the sol and gel stages in manufacture, and which is therefore amor- 
phous and highly porous in structure. The granular silica gel used in 
air conditioning units is equivalent in texture to 6 to 14 mesh, and has 
the appearance of colorless, semi-transparent sand, although its 
specific gravity is less than that of crystalline quartz because of its 
porosity. The pores represent the water of hydration which was re- 
moved when the material was converted from a gel to a solid. They 
are smaller than the wavelength of light, and are invisible under the 
ultra-microscope . 

This structure gives silica gel remarkable properties. The intense 
force of the resultant capillarity enables the granules to adsorb vapors 
within the gel granules, thus making it possible to separate vapors and 
imperfect gases from air and other perfect gases. The granules will 
take up from 30 to 50 per cent of their own weight of water vapor 
depending upon the conditions, without swelling or becoming exter- 



April, 1932] 



SILICA GEL AIR CONDITIONING 



473 



nally moist. One must think in molecular dimensions to realize that 
one cubic inch of the material has an internal surface of over one acre, 
and that when ground to the fineness of flour, only two per cent of its 
internal structure is destroyed, absorption tests proving that it retains 
98 per cent of its original adsorptive capacity. 

When the gel has taken up its useful load of vapor, it is readily re- 
activated by heat, which drives off the adsorbed vapors; after 
reactivation it is ready to be used again. The action in both cases is 
the purely mechanical action of capillarity working against vapor 



r 



40 



X 



. 

6- 



50 

IN GEL 



O fO 30 40 

7 CONCENTRATION OF 
FIG. 1. Curves showing performance of silica gel in dehydrating air of viiri- 
ous initial saturations at 25 C. and at atmospheric pressure. A, air at 100 
per cent saturation; B, 60 per cent; C, 40 per cent; D, 20 per cent. 

tension. No chemical reaction is involved, so that there is no deterio- 
ration of the gel; and the cycle of operations can be indefinitely 
repeated. 

The capacity and efficiency of silica gel as an adsorbent is dependent 
on a number of factors : the composition, the temperature, and the 
pressure of the gas-vapor mixture, the partial pressure of the vapor to 
be treated, the rate of flow per unit weight of gel, and whether the 
treatment is single or multi-stage, approaching adiabetic or isothermal 
operation. 

It is impossible in the limits of this paper to give exhaustive data for 



474 E. C. HOLDEN [j. S. M. P. E. 

all conditions. Fig. 1 shows its performance when dehydrating air of 
various initial saturations at 25C. and at atmospheric pressure. To 
determine the useful gel saturation in operation, the residual satura- 
tion of the gel of from 4 to 7 per cent should be deducted from the total 
saturations shown. 

In practice the efficiency of adsorption may be made to exceed 99 
per cent, depending on the type and size of the unit used, by keeping 
the operating cycle within the "break-point" limit. Some installa- 
tions are guaranteed to deliver air at 30F. dewpoint. By treating 
a regulated fraction of a total flow of gas-vapor mixture, any desired 
saturation can be produced; and the high efficiency of adsorption 
makes it possible to treat a minimum of the total circulation. 

It is because of the power of silica gel to maintain a vacuum greater 
than 29 inches of mercury in a vapor system that it finds its applica- 
tion in refrigeration, the adsorber taking the place of the compressor. 

Silica gel shows a similar selectivity for liquids, due to the character 
of the internal gel surfaces and the differences in surface tension of 
various liquids, thus making possible the separation and purification 
of hydrocarbons and other liquids; however, this class of applications 
is not of direct interest in the present paper. 

CONDITIONING AIR FOR FILM DRYING 

There are two practical stages in drying film, first: the removal of 
the excess surface water by the compressed air "squeegee," and, 
second, the removal of the water contained in the swollen films down 
to approximately 15 per cent residual hydration, required to keep film 
flexible and durable. These requirements are quite distinct and 
should be considered separately. 

In the preliminary stage, blowing off the moisture on the wet film by 
compressed air at the squeegee, the air should be clean, oil-free, and 
anhydrous, but the treatment actually used to condition the com- 
pressed air, so far as the writer knows, is to pass it through the usual 
compressed air receiver followed by some form of strainer; or at best, 
a simple type of air filter, as described by Crabtree and Ives, 1 for re- 
moving the condensate of compression and entrained compressor 
cylinder oil. No practical mechanical filter is 100 per cent efficient, 
and as it cannot remove vapor, a decrease of temperature between 
the separators and the squeegee causes further condensation of water 
and oil vapor; and finally, as the air expands at the nozzle and thus 
becomes chilled, more vapor will condense. 



April, 1932] 



SILICA GEL AIR CONDITIONING 



CONDITIONING COMPRESSED AIR 



475 



When air is compressed, some of the compressor lubricant^ is 
mechanically entrained in the air flow as a fine mist, and some of 
it, even though the highest test oil be used, is partially cracked and 
vaporized by the heat of compression. If efficient separating re- 
ceivers and mechanical filters be used after the compression, a large 
part, but not all, of the liquid oil-mist and water-condensate of 
compression settles out, although none of the true vapor of the oil or 
water is removed, these vapors passing on''and condensing later in the 
line, especially at the discharge, due to cooling on re-expansion. 



STEAM & WATER 
CONNECTIONS 




FIG. 2. Compressed gas dehydration unit. 

This can be entirely prevented and the air can be dried to a dew- 
point below any possible expansion temperature, and all oil vapor as 
well as oil-mist will be removed by inserting a silica gel pressure type 
adsorbing unit anywhere in the compressed air line following the 
receiver. The air passes through a bed of silica gel which adsorbs 
both the oil and water vapors and returns practically anhydrous, 
clean air to the line. Such air delivered at an effective pressure 
through the nozzle at the "squeegee" should do more than merely 
blow off the excess water; it should deliver uniformly clean film and 
appreciably reduce the duty required of the drying cabinets. 



476 



E. C. HOLDEN 



[J. S. M. P. E. 



Fig. 2 shows one type of small compressed air or gas silica gel 
dehydrator. 

CONDITIONING AIR FOR FILM DRYING CABINETS 

Inasmuch as films are made of permeable organic material, they 
will distort and lose their durability just as timber warps when 
improperly seasoned, and drying requirements cannot be figured a 

ATMOSPHERIC VENT 
Z70 CFTM. 



t 



WET 



FILM 



DRYING 



CABINET 




730C.RM. 



1000 CF.M. 
65-50 y.- 




tlOC.FM. 



SILICA 
DEHYDRATOR 



FIG. 3. Silica gel film drying; schematic diagram 1000 cu. ft. 
per minute circulated; 10 pounds of water per hour evaporated; 
volumes not corrected for temperature. 

priori as can evaporation from metal surfaces, but must be determined 
by experience. 

It is not, therefore, in the province of this paper to decide, or even to 
offer, an opinion as to what are the ideal conditions for treating either 
positive or negative films, or how much an anhydrous squeegee that 
has not heretofore been available to the industry, may hasten and 
simplify the subsequent drying. This can so easily be done, however, 
that it would seem worth proving. 



April, 1932] 



SILICA GEL AIR CONDITIONING 



477 



The air conditioning system, now in common use in processing film, 
of spray cooling or refrigerating to remove some of the water vapor, 
or in winter, of spray humidifying followed by reheating, treats the 
whole air stream, the used wet air being blown to waste. 

The silica gel adsorption system, owing to its ability to deliver 
practically anhydrous air, treats only a fraction of the air circulated, 
this fraction having an excess moisture capacity corresponding to the 
quantity of water being removed from the film. The whole air 




FIG. 4. Unit for treating air or other gases continuously at low 
pressure. 

stream with its dried fraction can then be returned to the cabinets in a 
closed circuit. The absorbing operation releases the heat of adsorp- 
tion, which varies up to one-third more than the latent heat of the 
water removed, so that any additional heat requirement is reduced or 
eliminated, and the closed circuit and special filters and the gel bed 
assure perfectly clean air and a complete control of temperature and 
relative humidity. 



478 E. C. HOLDEN 

It does not seem logical to have to add water to a drying unit 
With the adsorption system, if it be required to increase the humidity, 
the hydrometric control automatically slows or stops the adsorber 
operation and throttles the waste blow-off, so that the moisture taken 
from the film itself quickly builds up the humidity to the desired 
point, when the control again automatically regulates the adsorber to 
maintain it, and the necessary output of wet air is discharged through 
the relief valve. 

As an example of how this works quantitatively, the flow-sheet, 
Fig. 3, is given, based for convenience on the circulation of 1000 cubic 
feet per minute, assuming that 10 pounds of water per hour are to be 
removed. It is to be noted that the dehydrator would operate only 
when the atmosphere contains more than 90 grains of water vapor 
per pound of dry air, if that is the desired entering humidity. 

It is apparent from the practice followed in many film laboratories 
that the desired absolute humidity of the air entering the cabinets is 
higher than the average absolute humidity of the atmosphere, and 
that, therefore, the normal pretreatment required for fresh air enter- 
ing the drying cabinets is humidification rather than dehumidification. 
Whenever the atmospheric humidity exceeds the allowable humidity 
of the air entering the cabinet, a drying unit will be useful for main- 
taining the drying capacity without increasing the temperature of 
operation. This means, however, that predrying is necessary only in 
humid summer weather when the drying unit would be a convenient 
auxiliary for maintaining production regardless of the weather. 

The type of unit required is shown in Fig. 4, and consists of an air 
filter with two single-stage bed adsorbers operating alternately, 
adsorbing and activating, thus being capable of continuous 24-hour 
production. 

REFERENCE 

1 CRABTREE, J. I., AND IVES, C.E.: "A Pneumatic Film Squeegee." Trans. Soc. 
Mot. Pict. Eng., XI (Aug., 1927), No. 30, p. 270. 



MEASUREMENTS WITH A REVERBERATION METER* 
V. L. CHRISLER AND W. F. SNYDER** 



Summary. A description is given of apparatus with which the rate of decay of 
sound energy in a room may be measured. A loud speaker is used as a source of 
sound. When the sound reaches a steady state the loud speaker circuit is opened 
and at the same time a timer is started. When the sound energy has decayed to some 
definite value the timer is automatically stopped. If made in a portable form this 
equipment may be used to study the acoustical properties of auditoriums. Attention 
is called to the errors which may occur in these measurements. 

With the advent of the talking picture, the determination of the 
sound absorption values of various materials has become of consider- 
able importance. The original method of measuring the coefficients 
of these materials is due to W. C. Sabine, and requires the use of a 
reverberation room and rather large samples of material. The 
inconvenience of this led to attempts to find a method which would 
permit the use of smaller samples. 

One of these attempts, known as the tube method, is shown in Fig. 1. 
A mathematical formula can be derived showing that the sound 
absorption coefficient of the sample placed at the end of the tube can 
be computed if the relative values of the amplitude at the maximum 
and minimum points of the standing wave system in the tube are 
measured. Unfortunately, the results obtained in this manner are 
not in agreement with those obtained in actual installations. For 
this reason the method has been abandoned. At the present time it 
is necessary to adhere to the original reverberation method to obtain 
results which can be depended upon in actual installations. 

Figs. 2 and 3 show a plan and cross-section of the reverberation 
room at the Bureau of Standards. To obtain satisfactory measure- 
ments it is absolutely essential that all external noise should be 
eliminated. The outer walls and roof have therefore been con- 

* Presented at the Spring, 1931, Meeting at Hollywood, Calif. Publication 
approved by the Director of the Bureau of Standards. 

** Bureau of Standards. U. S. Department of Commerce, Washington, D. C. 

479 



480 



V. L. CHRISLER AND W. F. SNYDER [j. s. M. P. E. 



structed so that they are unconnected with the inner walls and ceiling. 
Due to this construction outside noises are seldom heard. 

Fig. 4 shows an interior view of the reverberation room with a 
sample of material laid on the floor, and Fig. 5 shows the position of 
the observer while measuring the absorption of an audience. 

To make measurements in this manner requires a trained observer. 
The method is very tedious as approximately one thousand observa- 
tions are required in order to obtain satisfactory values of the absorp- 
tion coefficients of a sample at six frequencies. To eliminate the 
personal error of the observer and to make measurements more quickly 





Loud Speaker 



To Amplifier 



g Surface \ 



FIG. 1. Diagrammatic scheme of the tube method of measuring 
sound absorption coefficients. 



and more accurately, considerable work has been done at the Bureau 
of Standards, as well as at other laboratories, to develop a method in 
which all measurements are made with some instrument. 

The first attempt was by use of an oscillograph. 1 As the sound 
waves decay in a very irregular manner in most cases, it is desirable to 
take the average of a number of measurements in computing the 
results. Figs. 6 and 7 show the irregular way in which the sound may 
decay after the source has been cut off. Fig. 6 is for a frequency of 
128 cycles and Fig. 7 for a frequency of 512 cycles. If enough records 
are taken at each frequency and the measurements averaged, satis- 
factory results can be obtained, but this requires too much work. 



April, 1932] 



REVERBERATION METER 



481 



t 




O I 2 S 4 S Ft. 



FIG. 2. Plan view of the reverberation room at the Bureau of Standards. 



SECTION A-A 



FIG. 3. Cross-section of the reverberation room at the Bureau of 

Standards. 



482 V. L. CHRISLER AND W. F. SNYDER [J. S. M. P. E. 

The most satisfactory arrangement 2 that has been tried is repre- 
sented schematically in Fig. 8. The source of sound is a loud speaker 
supplied with an alternating current of the desired frequency from a 
suitable oscillator and amplifier. The sound is picked up by a 
condenser microphone, and is then amplified. The purpose of the 
attenuator will appear from the following text. It is desired to call 
attention to the section of the circuit following the amplifier marked 




FIG. 4. Interior view of the reverberation room at the Bureau of Standards, 
showing a sample of acoustical material on the floor. 

"tube relay," which consists of a rectifier tube followed by a stage of 
direct current amplification. The circuit is shown in Fig. 9. 

These tubes are connected in such a manner that, after the alternat- 
ing potential applied to the first tube decreases to a definite value, a 
very small additional decrease causes a relatively large increase of the 
plate current of the last tube. This has been accomplished by util- 
izing a "freak" characteristic of the first tube. Fig. 10 shows the 



April, 1932] 



REVERBERATION METER 



483 



static characteristic of this tube and gives the variation of the screen- 
grid current and the plate current with the grid potential when a 




FIG. 5. View showing the position of the observer while measuring the acous- 
tical absorption of an audience. 

definite screen-grid potential is used. To produce such a characteristic 
only a limited range of screen-grid potentials can be used. If the tube 




FIG. 6. Oscillogram showing the decay of sound after the source has 
been cut off; 128 cycles. 

is biased so as to obtain rectification at the upper end of the curve, 
advantage can be taken of the abrupt change in plate current with a 
very small change in grid voltage. 



484 



V. L. CHRISLER AND W. F. SNYDER [j. s. M. P. E. 



Fig. 11 shows the modified plate current when an alternating 
voltage is applied to the grid of this tube, and the corresponding 
change in plate current in the last tube. 

The sudden increase of the plate current of the last tube operates a 
relay which stops the timer. As the timer is started automatically 
when the loud speaker current is broken, this device gives the time 
required for the sound to decay to some level determined by the 
amount of amplification used. 




FIG. 7. Oscillogram showing the decay of sound after the source has 
been cut off; 512 cycles. 

By using an attenuator in the amplifier circuit the time required 
for the sound to decay to different levels can be determined. In this 
way a decay curve can be obtained similar to that obtained in calibrat- 
ing a room by the ear method, which uses different intensities, the 
ratios of which are known. 

There is one marked difference between these two methods. In 

CONPCNSER 
MICROPHONE 





3 STAGE 

RESISTANCE -COUPLED 

AMPLIFIER 

95 db gain 



FIG. 8. 



Schematic arrangement used at the Bureau of Standards for 
making acoustical measurements. 



the ear method we start at different intensity levels and end always at 
the same lower level, which is the threshold for the ear of the observer. 
In the instrumental method we always start at the same intensity 
level and end at arbitrary thresholds whose ratios are known. Fig. 
12 shows a curve thus obtained. It will be observed that the points 
fit a straight line very closely. The points on the curve are not the 
results of single measurements but are each the average of ten mea- 



April, 1932] 



REVERBERATION METER 



485 



surements. With a timer which adds, several measurements can be 
taken rapidly and the average obtained, thus eliminating the un- 
certainties of a single measurement. 

UY224 .I.I.I.I.M.I.I.I UY224 




FIG. 9. Circuit diagram of the vacuum tube relay. 



To timer 



The slope of this line gives the rate of decay of the sound energy. 
From this slope the reverberation time may be computed, as reverbera- 
tion time has been defined as the time required for sound to decay 



51.0 

1.9 

.S 

I- 8 
5.7 

(B .6 

5 

ft 

A 

.3 



Radiotron UY-234 

Plate Voltage- 90 v. 

Screen-Grid Voltage -34>5v. 

Filament Vo/taae-Z^v. 
Plate Resistance -^megohm 
Grid Bias to Cathode 
and neqatt^ Fi lament. 




Grid Bias -volts 

FIG. 10. Static characteristic of the recti- 
fier tube of Fig. 9. 

sixty decibels. Knowing the reverberation time, the total absorption 
of a room can be computed either by Sabine's formula or Eyring's 
general reverberation equation, as may be desired. 



486 



V. L. CHRISLER AND W. F. SNYDER [j. s. M. P. E. 



This arrangement gives a satisfactory instrumental method of 
measuring sound absorption, and also a method of determining the 
reverberation time of any room. 

Satisfactory equipment for making these measurements does not 
solve all the difficulties of making such measurements. If an ac- 
curacy of not greater than ten per cent is desired in the total absorp- 
tion, most of the difficulties vanish; but when greater accuracy is 
desired several precautions must be taken to obtain a uniform distribu- 
tion of sound energy in the room where the measurements are made. 

To obtain such a distribution a band of frequencies was used at 



pe 



.I/ 
.13 

07 


(/* 


'224 R 
P/ate 
^creen-G 
niame 
Vafe Res 
Grid h 
rid Has 
neyat 


ectifier ~i 
voltage 
rid votta 
nt vo/fa 
Distance 
las -1.4-v 
^o Ccrfhot 
Ve FIla^ 


r rigger D 
90 v. 
ge -34.5 
ge-2.Sv. 
- .5 mega! 
neg. 


evice 


\ [ 


P 


V. 
int 


j) 


G 


ie and 
nent 


7 










/ 




\ 




^ 


^ 






j 




Relay Co 


rrent 


. 


/^ 


* 



3.0 



20 



I 

D 

u 

1- 

<u 
QC 



1.0 



FIG. 11. 



25 2. t.5 I 3 Q 

A.C. Potential -applied to Grid 
Response characteristic of the rectifier tube of Fig. 9. 



first ; but later work has shown that this is undesirable, as beat notes 
may occur, which appreciably alter the result. The source of sound 
is in constant motion, this motion aiding materially in producing a 
uniform distribution. At the higher frequencies it was thought that 
this motion would be unnecessary, but it was found that the apparent 
rate of decay of a sound might vary ten per cent when both the source 
of sound and the microphone were stationary. This random varia- 
tion rarely exceeds two per cent when the source is in motion. 

When making measurements in a reverberation room it is possible 
to take these precautions, but in studying the rate of decay in a 
theater or auditorium, it becomes more difficult. 



April, 1932] 



REVERBERATION METER 



487 



To make an intelligent application of acoustical material in a 
theater it is believed that equipment such as described here, or the 
reverberation meter as developed by the Bell Telephone Laboratories, 
should be used to study typical auditoriums and to learn more about 
sound distribution and rates of decay in different portions of the room. 



44 

a/; 




























/ 


* 


























./ 


/ 


























/ 




























/ 










I 

-28 

c 
c 

12 
4 




















/ 




























x 




























x 




























x" 




























x 


Q 4Vx2.3(log K) E 2 -log lo E < ) 












x 




v(t 2 -t ( ) 

/ . \ 








x 






Absorpti( 


5n= 




__^-_ 


cycles per sec. 

1 1 1 1 






x 






1 


^1^ 

1 1 












^_ 



5 10 15 

Time in seconds 

FIG. 12. Decay curve obtained by making measurements of a 
room, starting at the same intensity level and ending at arbitrary 
thresholds whose ratios are known. Each point represents the 
average of 10 measurements. 

This study should be made at all frequencies so as to aid in determin- 
ing the most desirable characteristics of a sound absorbing material 
and the locations in which such material should be applied. 

REFERENCES 

1 MEYER, E., AND JUST, PAUL: "Zur Messung von Nachhalldauer und Schall- 
absorption," Elek. Nach. Tech., 5 (1928), p. 293. CHRISLER, V. L., AND SNYDER, 
W. F.: "The Measurement of Sound Absorption," Bureau of Standards Jour, of 
Research, 5 (Oct., 1930), p. 957. 

2 MEYER, E. : "Automatic Reverberation Measurement," Zeit.f. Tech. Physik., 
II (1930), No. 7, p. 253. STRUTT, M. J. O.: "Automatic Reverberation Measur- 
ing Instrument," Elek. Nach. Tech., 7 (July, 1930), p. 280. WENTE, E. C., AND 
BEDELL, E. H.: "A Chronographic Method of Measuring Reverberation Time," 
/. Acoust. Soc. ofAmer., I (April, 1930), No. 3, p. 422. 



16 MM. SOUND FILM DIMENSIONS* 
RUSSELL P. MAY** 

Summary. A method is set forth for the derivation of dimensions and locations 
of the final projection print and all camera, printer, and recording apertures, con- 
sideration being made for film weave, shrinkage, and mechanical tolerances in the 
apparatus involved in producing and projecting the film. 

Two methods of producing films are considered: (a) Projection positive print 
made from a 16 mm. dupe negativz by continuous contact printing, where the dupe 
negative is made by optical reduction of the 35 mm. picture from a master positive 
and the sound re-recorded from a 35 mm. sound track, and (b) production of the pro- 
jection positive print from a 35 mm. picture negative by optical reduction and re- 
recording of the sound from the 35 mm. sound film directly to the final 16 mm. 
positive. The method provides for modification of these processes so that any com- 
bination can be used. 

Motion pictures in the home have in the past three or four years 
enjoyed a slow but steadily increasing popularity and utility. One 
witnesses frequently at the beaches and other resorts, amateur 
cinematographers with their cameras making pictorial records of 
their children's and friends' animations. Each year these experiences 
have become more frequent and now it is not unusual to be enter- 
tained, during an evening's visit, with motion pictures whose prin- 
cipals are your own friends and acquaintances. At the motion pic- 
ture counters of photographic supply houses we see interested people 
discussing cameras, projectors, lenses, etc., or leaving cine-films to be 
developed. By these activities it is not difficult to conclude that the 
public is, in part at least, cinette-minded. 

Development of motion picture equipment in all its branches has 
advanced in amazing strides since the introduction of sound, and 
paralleling this, the home sound movie has likewise been developed. 
Numerous devices have already made their appearance on the market. 
Thus far, they all employ disk type sound records driven synchro- 
nously with the film, but sound projectors utilizing sound on film are 
soon to make their debut. 

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. 
** RCA Victor Co., Camden, N. J. 
488 



16 MM. SOUND FILM DIMENSIONS 489 

In order to present this subject clearly, it is desirable to review the 
difficulties encountered in the early attempts at interchangeability of 
films made by the various producers of sound films of the variable 
width and variable density types. Innumerable cases of variations 
of locations of sound track, recorder, and reproducer optical systems 
contributed to endless difficulties in attempts to arrive at universal 
operation and satisfactory performance of reproducing equipments. 
Augmenting these difficulties another source of trouble arose due to 
inherent requirements of the variable width and variable density 
type sound records, the former requiring that the scanning slit fully 
cover the record at all times, allowance being made for variations 
that might be introduced during the production of the projection 
print or in the projector. Should the end of the scanning slit fall 
within the boundaries of the record, the peaks would be "chopped 
off," thereby introducing distortion in the reproduced sound. It is 
therefore evident that the sound track width should be somewhat 
less than the length of the scanning slit. 

In the case of the variable density type record, the opposite re- 
quirement, that the scanning slit should at no time fall outside 
the boundaries of the sound track, applies. Should this occur, 
noise might be introduced by either the sprocket holes or the 
picture. 

Thus we see that if a universal scanning slit is to be used in pro- 
jectors, the first-mentioned record must be narrower than the latter 
and the locations of the records and the scanning slit must be held to 
close limits. 

It is needless to dwell on the desirability of preventing a recurrence 
of past difficulties with the advent of home sound motion pictures. 
Surely no word of explanation is needed to point out the importance 
of standardized film, recording and reproducing slit dimensions, 
when considering the potential home and industrial fields for this 
class of equipment and the production of apparatus and films, the 
success of which depends wholly upon interchangeability of films in 
the projection equipment. 

With this point in view, this paper sets forth a method for the 
determination of film dimensions taking into account the variations 
experienced by the films from the making of the original 16 mm. 
negative to the projection of the positive print. This method has 
been followed in arriving at the projection print dimensions, as well 
as the projector sound and picture aperture dimensions. 



490 



RUSSELL P. MAY 



[J. S. M. P. E. 



The diagram in Fig. 1 shows all the probable steps in the production 
of the positive film from a "dupe" sound and picture negative and 
the extreme variations considered. It may be advantageous for 

I 



Jo ^f 




be 

il 



IE 



I! 



s 

o ^ 

if 

Bg 

1 

a*a 
a D 

i-s 
'.I 

</) C|_| 

a, o 

3 8 



11 



fitf 

.2 
. VJ3 

'".S 
d 5 



various reasons to produce projection prints by direct reduction of 
the picture from a 35 mm. negative and re-recording of the sound 
from the 35 mm. film direct to the projection positive. Fig. 2 shows 
the various steps involved in this procedure. It will be noted that 



April, 1932] 16 MM. SOUND FlLM DIMENSIONS 

Jo fp piCiq |, 



491 




492 



RUSSELL P. MAY 



[J. S. M. P. E. 



the sound and picture apertures in both diagrams are the same. This 
was done in order that the processing equipment might be uni- 
versally adaptable to both methods. The latter method involves fewer 
steps, which will increase the margin of safety. 

The variations used in the original calculations were twice those 
shown in the diagrams and were based on the practical limits of the 
machines to which they apply. A film 0.660 inch wide resulted, 
which did not meet the requirement that the sound and picture be 
adapted to a 16 mm. film. Examination of the various steps shown 
in Fig. 1 discloses that each factor varies, within certain limits, in- 




FIG. 3. Light slit dimensions and loca- 
tion for making dupe negatives on 16 mm. 
sound recorder. The diagram shows the 
film in the recording position with the 
emulsion facing the recording light. 



dependency of any other. By applying the principle of probability 
to the distribution between these limits, the likelihood that all the 
variations will simultaneously occur in the same direction is so remote 
that it falls within the bounds of safety to reduce the original limits 
to one-half those shown in the diagram. By so doing, we are en- 
abled to use standard 16 mm. film to carry the sound track in addition 
to facilitating the projection of present amateur films. 

The projection print will be a 16 mm. film having standard perfora- 
tions along one edge only. Eliminating the perforations along the op- 
posite edge provides space for the sound track. 



April, 1932] 



16 MM. SOUND FILM DIMENSIONS 



493 



The requirements to be fulfilled are as follows: 

(1) 16 Mm. film having one row of sprocket holes. 

(2) Standard projector picture aperture. 

(3) Projector picture aperture within the film image at all times. 

(4) Clear space between picture and sound for a film supporting shoe in 

recording, printing, and reproducing equipment. 



.1186" 



3JTIT 



r 



i 



P .3981 

.5175 - 



a 

Emalsion Docun 



FIG. 4. Printing aperture dimen- 
sions and locations for making direct 
reductions to 16 mm. dupe negatives 
on optical reduction printer. Diagram 
shows film in printing position with the 
emulsion facing the printing light. 



(5) 60-Mil recording slit. 

(6) Sound reproducing slit to cover track at all times with ends riding 

in opaque stripes adjacent to the track. 

(7) Transparent space at sound edge of film to prevent peeling off of 

emulsion. 

Starting at the top of the diagram, Fig. 1, the solid areas represent 
aperture plates and the clear spaces the apertures through which 
the sound recording and picture printing lights pass as indicated at 



494 



RUSSELL P. MAY 



[J. S. M. P. E. 



1, 2, 3, and 4. The lines thus numbered will be referred to as the 
"principal lines." The departure of latent or developed images from 
these starting points is shown by the parallel lines, diverging at each 
step, for the reasons indicated. Lines diverging to the left have 
been designated a and those to the right b, in each instance. These 
lines denote extreme limits only, and when considering picture or 

.6269" 



Em ul si on Up 




.00 55" 



FIG. 5. Sound track and printer dimensions and loca- 
tions; 16 mm. dupe negative (developed, to 0.5 per cent 
shrunk). The sound record leads the picture by twenty- 
one frames or 6.3 inches. 

sound track widths, the total distance between lines having similar 
designations is determined by addition. 

The divergence continues in a regular manner in both directions 
down through steps A , B, and C. Variations shown in step D deal 
with dupe negative shrinkage and therefore are in one direction only, 
and since shrinkage of the dupe negative can result only in displace- 
ment to the right, the a lines continue down from C to D without 
change. The dupe negative is guided in the continuous printer 
by the sprocket holes, which in the diagram are represented by the 



April, 1932] 



16 MM. SOUND FILM DIMENSIONS 



495 



vertical line at the right-hand edge, and causes the shifting of the 
film image, due to the shrinkage, to occur in the direction shown. 
The line, 4b, in this step does not show a shrinkage offset This was 
purposely omitted as it only amounts to 0.00001 inch and can be 
neglected. 

Step E shows the film image positions after printing the dupe nega- 
tive in the continuous printer, whose variations are as indicated. 
The solid areas indicate light-shields built in the printer. The solid 
area near the center is a light-shield and film support to hold the dupe 
negative and positive films in contact. The space between this and 
the shield on the left represents the opening of the sound printing 
light aperture. It will be noted that a space exists between the 

Dupe Neq>+ive cui4)i 
emulsion up. 




q sprocket 
fo serve &s film 



Sound prm-finq 
liqhl &per4ure^ 

.0060 



ide. 



Positive film cui4h 
emulsion docan. 



Picfure 



U-.005L_Nofe=-Heiq)rf of pnVHnq 
1 . *per+ urc -UTsu^ liqh+ * 
film speed. 



Ucjh+ 5Vneld frv-nd' -*j J-^.o7O 

Film 5upporf 

FIG. 6. Printing apertures, dimensions and locations; 16 mm. continuous 
contact printer. Diagram shows negative and positive films in printing po- 
sition, emulsion facing printing lights. 

maximum sound track limits and the ends of this aperture. It is 
through these clear spaces that the light of the sound printer passes 
to print the black stripes at either side of the sound track. The 
space occupied by the light-shields will receive no light, and therefore 
will result in transparent stripes on the developed positive film. 
The right-hand edge of the center light-shield coincides with the 
projected line 3a, and represents the left end of the picture printing 
light aperture and, as shown, coincides with the picture negative in 
its extreme left position. When the picture negative assumes an 
extreme right position a clear space of 0.006 inch will occur between 



496 



RUSSELL P. MAY 



[J. S. M. P. E. 



it and the end of the aperture, and will show up as a black stripe on 
the positive. The opposite end of this aperture is located at the 
projection of line 4b at step E. Addition of the dimensions involved 
results in an aperture 0.4027 inch long, the end of which is located 
0.002 inch from the sprocket holes. Under these conditions we may 
be sure that the picture will never be cut off in the printer, and the 
sound track will at all times be bordered with black stripes. 




FIG. 7. Sound track and picture dimensions 
and locations; 16 mm. positive (developed, to 1.5 
per cent shrunk) made from dupe negative. 

Step F deals with shrinkage of the positive film, and shows that 
the image shifts to the left. This is due to the fact that during pro- 
jection the film is guided by the edge of the film adjacent to the sound 
track and, therefore, any motion of the film image due to shrinkage 
will be in this direction. 

The opaque light-shields shown in step E introduce a new set of 
secondary lines, 5, 6, and 7, which must be considered in all subse- 
quent^ steps. It will be noted that these lines undergo the same 



April, 1932] 



16 MM. SOUND FILM DIMENSIONS 



497 



divergence as the principal lines. Line 5a is shown dotted, as it 
falls without the boundaries of the reproducer aperture and is un- 
important; the same holds true for line 7 a. Line 5b, however, con- 
verges toward principal line la, and 6a toward principal line 2b, at 
each succeeding step. The divergence of the principal lines continues 
through steps G and H for the reason noted, finally meeting the second- 
ary converging lines. Attention is directed to the fact that lines 
2b and 6a do not actually intersect at step H. This would ordinarily 
occur, but in this particular case a modification was necessary to 



a 


^ 


3 


'5 


> 


t-y 


/ 


w 




J 


> .0119 H 


, 


.0066 






' .0599 


3 


* 


^ .Ol*ffc 






-. .0745 


; 


. 




j 


*/ 




: 


& 




:> 


\ 




Emulsion Up 


j 





FIG. 8. Light slit dimensions and location 
for making direct re-recordings on 16 mm. 
sound recorder; diagram shows film in record- 
ing position with emulsion facing the record- 
ing light. 

adapt the reproducing aperture to film made either by printing or 
by direct re-recording. These points of intersection define the limits 
of the projector sound and picture apertures. The shoe shown be- 
tween the two apertures serves as a support for the film and rides 
in the clear space indicated by the lines converging from the center 
printer shield. The dimension 0.0228 inch defines the limits to which 
the edges of the clear space can move toward each other and not its 
actual width, thus assuring that the shoe in the position shown will at 
all times ride on an emulsionless surface. By using the intersections 
of the progressive variations as locating points for the aperture, we 
are assured of proper registration with the film images. 



498 



RUSSELL P. MAY 



[J. S. M. P. E, 



In the construction of the diagram, the width of the film from the 
inside edge of the sprocket holes to the opposite edge of the film, 
0.5215 inch, is used as the working basis, and satisfies requirement 
No. 1. 

Having predetermined all the factors which should be satisfied, 



! 

a o- 

c 



D 



a 



1I1FE 



* .380'^ -- H 
.5048 



a 



Emalsion Up 



-*|.l085h 



Vieca fVcinq proecfion 
light ^nd. shocjainq film 
in projection position. 



.O711 
.OIO' 



FIG. 9. Picture aperture and scanning line dimensions and locations; 
16 mm. projector. Diagram shows film in projection position facing 
projection light. 



the diagram was laid out completely as shown, minus the dimen- 
sions. Requirement No. 2 was met by locating the 0.380 inch pro- 
jector picture aperture at the center of the film, having the 0.0167 inch 
distance to the sprocket holes. By constructing lines 4a and 4b 
so that they are located between the picture apertures and sprocket 
holes, requirement No. 3 is fulfilled. The dimensions shown are 
derived from the 0.0167 inch spacing and the predetermined varia- 



April, 1932J 



16 MM. SOUND FILM DIMENSIONS 



499 



tions indicated by the verticaj figures. The point of coincidence at 4 
determines the right-hand edge of the printer aperture. Point 3 is 
located by similar treatment. 

A shoe width of not less than 0.020 inch was considered desirable, 
and after an initial calculation 0.0228 inch was chosen as it facilitated 
the use of standard gauge sheet metal for the light-shield. Require- 
ment No. 4 is thereby satisfied. 




FIG. 10. Sound track and picture dimensions 
and locations for 16 mm. positive (developed, to 
1.5 per cent shrunk) made by direct picture reduc- 
tion and sound re-recording. 

The position of 2b at step H locates the right-hand end of the sound 
reproducing aperture or slit. Requirement No. 5 is met by making 
the distance between lines 1 and 2 at step A 0.060 inch and the posi- 
tion of la at step H satisfies requirement No. 6; working up from 
this point to secondary line 5 results in a distance from the edge of 
the film of 0.008 inch, which satisfies requirement No. 7. Fig. 2 
shows the method as applied to 16 mm. sound films made by 
direct re-recording of sound and reduction of picture from 35 mm. 



500 RUSSELL P. MAY [j. s. M. P. E. 

dupe inegative. The aperture dimensions and location are identical 
to those shown in Fig. 1. 

The two methods shown in the diagrams provide a flexible arrange- 
ment whereby 16 mm. projection prints may be made using either 
method or any combination thereof. In other words, in addition 
to the methods shown in the diagrams the sound may be re-recorded 
and the picture printed from a 16 mm. dupe negative, or the sound 
may be printed from a 16 mm. dupe negative and the picture printed 
by optical reduction from a 35 mm. dupe negative. In any case, we 
may be assured, however they are made, that they will always 
register properly in the projector. 

With regard to sound tracks of the variable density type, it will be 
observed that adequate provision has been made to permit the 
scanning slit to fall within the track area if the methods set forth 
are followed, the additional width being provided by utilizing the 
space occupied by the black stripes in the variable width type of 
record. 

The accompanying drawings show in detail all aperture dimensions 
and locations, and maximum and minimum film dimensions as de- 
rived by the foregoing methods. 

DISCUSSION 

MR. MITCHELL: Judging from our experience on 35 mm. film, we ought to 
consider maximum shrinkage. The fact that 16 mm. film is made of safety stock, 
which shrinks more than standard stock, should also be considered. 

The elimination of one row of perforations will tend to obsolete a lot of 16 mm. 
film already in use both here and abroad. 

MR. MAY: We have considered the matter of obsolescence of the existing 
16 mm. equipment, and feel that it is not as important as it may seem, due to 
the fact that a picture made for sound requires sound with it in order to afford 
a satisfactory performance. 

We do not think it important to be able to show sound films on existing 16 mm. 
projectors, but we do feel that the projection of existing 16 mm. films made with 
amateur cameras should be capable of being projected in our new sound equip- 
ment. That is possible. As to shrinkages, we have allowed somewhat more than 
what we encounter in practice. One and one-half per cent is ample for the 
positive. 

MR. VICTOR: Is this paper intended as a proposal for a standard? 

MR. MAY: Yes. 

MR. VICTOR: Have you experimented with contact printing from 16 mm. 
negatives to positives, with the sound track? 

MR. MAY: We have, and are satisfied that it can be done commercially. 

MR. VICTOR: Is the emulsion fine enough? 

MR. MAY: I might add that it is not as good as we would like it to be. But 



April, 1932] 16 MM. SOUND FlLM DIMENSIONS 501 

it is our opinion that the projected picture and the sound quality are commercially 
satisfactory. 

MR. SPONABLE: I note that the distance between the sound records of the 
corresponding pictures is 6.3 inches. That is not a straight reduction from thirty- 
five to sixteen, is it? 

MR. MAY: No. The two distances that are used are for theater use. The 
distance used by RCA Photophone, Inc., is somewhat different from that used 
in the Movietone. If I am not mistaken, it is on the opposite side, too, is it not? 

MR. SPONABLE: It is my impression that the separation is fifteen and a half 
inches, for theater use. It seems to me that if the 16 mm. machine were properly 
designed, we could use a straight reduction from 35 mm. film. 

MR. MAY: That would be true if we were making it from a sound negative of 
which the dupe is already made. It might be more practicable to print first the 
sound and then the picture, in which case the dupe negative could be run through 
the two printers one after the other, in which case the sound lead would not be 
important. 

MR. SPONABLE: I wondered whether it was a case of not being able to design 
a projector that would give a displacement of four and three-quarter inches? 

MR. MAY: There would result little less than the present length of six inches. 
This is as short as we can make it and still get the various mechanisms in place. 
We have been using a single row of perforations for probably two and a half years 
in our development work, and have found that the wearing qualities of the film 
show no appreciable difference, whether one or two rows of perforations are used. 
In fact, we have run films to destruction, and in most cases cannot keep the splices 
in long enough to finish a film. I might add that twelve or fifteen hundred trips 
to the gate are not uncommon. And usually the emulsion and the surface of the 
film are damaged, rather than the sprocket holes. 

MR. RICHARDSON: How is it proposed to adapt the sound to the relatively 
tiny figures in the 16 mm. screen image? 

Using the relatively narrow 16 mm. sound track, is it possible to obtain the 
same quality of the sound that may be had from the wider track? 

MR. MAY : The picture can be made about as wide as five feet. When sitting 
in a small room, the angle at which the eye subtends the screen is not greatly 
different from the similar angle in the theater. In other words, sitting in a living 
room and watching a picture as large as can be projected from a 16 mm. film, 
one obtains the same illusion of size as he does in the theater. As to the sound, 
the sound track is only ten mils narrower than that used on 35 mm. film. The 
latter has a seventy-mil track, and we have used a sixty-mil track in our small 
film. The difference in size of track is only about sixteen per cent. The sound 
output is about the same as that obtained from a radio set comfortable room 
volume. 

MR. HICKMAN: This is not an attempt to form standards at any immature 
stage in the development of the art. The complaint has been made, in previous 
developments of motion picture engineering, that dimensional cooperation is 
absent at the start; after the job is done the manufacturers try to get together 
and find out how they can simplify matters. Now we have the advantage here of 
someone's thinking out clearly and carefully, at the very beginning, the factors 
that underly the situation; presenting them quite openly, so that what is being 



502 RUSSELL P. MAY 

thought by one powerful group will be known by all. I do not think that there 
is any question of establishing these standards at this time. 

Mr. Hardy asked me to announce that this paper was presented under full 
cognizance of the Standards Committee and is being considered by them. 

MR. EVANS: One of the reasons why the Standards Committee wanted 
to have this paper presented before the Society at this time was to discover 
what, if any, objections to it might be raised, so that the Committee would 
have as many facts before it as possible when it attempts to standardize 16 mm. 
pictures. 

To the Committee were presented several communications indicating quite 
different viewpoints on the subject. One communication advised that although 
we were going to have a 16 mm. sound track, it ought to be 20 mm. Another one 
stated that if we were going to have a 16 mm. sound track, two rows of perfora- 
tion should be used. Those were important differences. The Committee wel- 
comes free discussion of all such important factors, so that it will have all the 
various viewpoints before it. 

MR. COOK: After an experience of eight years in the Kodascope Libraries, 
I can assure you that a single row of perforations is ample to secure a much 
greater projection life of the film than is likely to be consistent with the obsoles- 
cence of the subject. The Bell and Howell machine has only one claw on one 
side; test strips have been run in this machine many hundreds of times without 
apparent deterioration of the edges of the perforations. 

MR. RICHARDSON: With a single line of perforation, would it not be necessary 
to adjust the tension carefully? 

MR. COOK: In the 16 mm. film a more accurate registration is possible with 
a single row of perforations than is likely in the 35 mm. film with two rows of 
perforations. There is no difference that the unaided eye is able to discover 
in the image projected from a film with a single row of perforations than from 
one with two rows. 



PROPOSED CHANGE IN THE PRESENT STANDARDS 
OF 35 MM. FILM PERFORATIONS* 

A. S. HO WELL AND J. A. DUBRAY** 

Summary. There are at the present time two standards of 3 5 mm. film perfora- 
tion, one known as the Bell & Howell perforation for negative films and the other the 
rectangular perforation for positive films, both of which have been approved and 
adopted by the Society of Motion Picture Engineers. Unfortunately, the use of these 
two standards introduces complications found detrimental in certain types of work, 
which indicate the advisability of having a single standard. 

It is felt that the rectangular style of perforation has advantages that it is de- 
sirable to retain. An alternative standard is proposed that will combine the ad- 
vantages of both the present styles, and which, at the same time, can be used on prac- 
tically all existing equipment without alteration of that equipment. Means are 
also suggested for shortening the transitional period of such a change-over. 

There are at the present time two standards of 35 mm. film perfora- 
tion, one known as the Bell & Howell perforation for negative films 
and the other the rectangular perforation for positive films. Both 
standards have been approved and adopted by the Society of Motion 
Picture Engineers, for reasons which are well known. 

It appears that while the decision to adopt two perforation stand- 
ards was originally taken with a view of reconciling commercial 
requirements and economic dictates, the adoption of a double stand- 
ard has now created an undesirable condition, especially in contact 
printing, process work, etc., involving exact superimposed regis- 
tration. 

While it is true that until now the motion picture industry has been 
able to get along with the two standards of perforation, it is equally 
true that this double standard creates a serious barrier to further 
technical advances of motion pictures. This barrier is perhaps only 
important at the present time to a limited number of motion picture 
technicians, but it is a barrier which will rapidly and seriously hamper 
the efforts of the industry toward further achievements. 

In this connection, it is interesting to review the discussion that 

* Presented at the Fall, 1931, Meeting at Swampscott, Mass. 
** Bell & Howell Co., Chicago, 111. 

503 



504 



A. S. HOWELL AND J. A. DUBRAY [J. S. M. P. E. 



occurred on the rectangular perforation at the time it was proposed 
by the Standards Committee. 1 At that time, the difficulties arising 
from the use of two standards of perforations were pointed out, 
and the situation today, with respect to the impossibility of securing 
satisfactory registration when the two sizes of perforations had to 
be used together, was forecast. 

In present practice where superimposed registration is required, 
as in composite photography; dupe negatives for lap dissolves; 
step printing with pilot control; color photography, in which two 
negatives are exposed simultaneously in contact or in superimposed 
relation, or two frames exposed in accurate relation to one another, 



I 
FIG. 1. 



Splice of two films, one with negative and one 
with rectangular perforations. 



and many other conditions where it may be desirable to use positive 
and negative stock together, the existence of two dissimilar perfora- 
tion standards is already showing ill effects and will introduce even 
more serious obstacles in the near future. Furthermore, the dual 
perforation standard may eventually become a cause of trouble in 
theater projection, with the practice of indiscriminately using both 
positive and negative perforations in release prints. When both 
types of film perforation are used indiscriminately in this manner, 
it is impracticable, if not impossible, to maintain good splice registra- 
tion even with the best facilities available. 

For instance, Fig. 1 shows a splice of two films, one with negative 
and one with rectangular perforations. In the shaded outlines, two 
conditions of the positioning of the splicing machine pilot pins, with 



April, 1932] CHANGE IN 35 MM. FlLM PERFORATIONS 505 

respect to the perforations, are shown, the pilot being of the same 
size and shape for both conditions. In the upper part of the drawing, 
the pilot pins are assumed to fit the negative perforation perfectly. 
These same pilot pins cannot, however, fill the rectangular perfora- 
tion, and the drawing plainly shows what the maximum error in 
registration may be. 

When two sizes of perforation are used, as it is quite impr